Upcoming Events

From Standardization to
The Battleship Approach to the Low Resource Entity Matching Problem
Prof. Avigdor Gal

January 11, 2024, 11:00-12:00
Campus city, Amir Building,
Conference Hall 508

Computer Science Colloquium - Democratising NLP: Overcoming Language and Domain Barriers in Low-Resource Environments

Learn More >>

Conferences funded by the center

June 20th, 2023
University of Haifa

The 10th edition of the Haifa conference on marine sciences

Learn More >>
Learn More >>

Related Events

23.5.2024 Thursday 12:00
Carmel campus - main building
600 pavilion

תערוכה המציגה חידושים ומחקרים מתקדמים בתחום הרובוטיקה והטכנולוגיות התת-ימיות, וכוללת מספר מוצגים ומסך אינטאקטיבי. התערוכה הינה פרי שיתוף פעולה של המעבדה של פרופ’ רועי דיאמנט עם המעבדה לרובוטיקה ימית באוניברסיטת זגרב, קרואטיה, במסגרת פרויקט מחקר של האיחוד האירופאי ושת״פ של משרד המדע הישראלי והקרואטי, ומוצגת גם במוזיאון טסלה למדע בזאגרב 

2-4.6.2024 Sunday-Thursday
ELMA Hotel
All the information and registration is here.
Please notice:
Abstract submission due 7.3.2024

Related Events

23.5.2024 Thursday 12:00
Carmel campus - main building
600 pavilion

תערוכה המציגה חידושים ומחקרים מתקדמים בתחום הרובוטיקה והטכנולוגיות התת-ימיות, וכוללת מספר מוצגים ומסך אינטאקטיבי. התערוכה הינה פרי שיתוף פעולה של המעבדה של פרופ’ רועי דיאמנט עם המעבדה לרובוטיקה ימית באוניברסיטת זגרב, קרואטיה, במסגרת פרויקט מחקר של האיחוד האירופאי ושת״פ של משרד המדע הישראלי והקרואטי, ומוצגת גם במוזיאון טסלה למדע בזאגרב 

2-4.6.2024 Sunday-Thursday
ELMA Hotel
All the information and registration is here.
Please notice:
Abstract submission due 7.3.2024

Past Events

From Standardization to Theory and Back

Liat Peterfreund (Hebrew University)

Graph databases are becoming increasingly popular due to their natural data modeling, making them useful in expressing connections that are harder to express in the relational model. Indeed, graph databases are used in a plethora of domains ranging from social to biological networks, and for various use-cases including fraud detection and investigating journalism. Since 2019, GQL (Graph Query Language) is being developed under the auspices of ISO as the new standard for querying graph databases, akin to SQL for relational databases. In this talk, I will present a researcher’s digest of GQL by describing its underlying theoretical model. I will demonstrate how we can use tools from formal language and automata theory to show the limitations of this new standard, which can hint at extensions for its next versions.

This talk is based on joint works with Nadime Francis, Amélie Gheerbrant, Paolo Guagliardo, Leonid Libkin, Victor Marsault , Wim Martens, Filip Murlak,  Alexandra Rogova, and Domagoj Vrgoç.

Causal Data Integration

Guy Kornowski

Weizmann Institute

The unprecedented empirical success of deep learning poses major theoretical challenges. Training modern neural networks requires solving large-scale optimization problems which are neither convex nor smooth, while classic wisdom also suggests that they are too complex to be able to generalize in a meaningful manner.

In this talk I’ll present some recent advances aiming at these issues, which reveal the surprisingly beneficial roles played by several factors such as large gradients, randomness, cheap updates, high dimensions, and training even after overfitting.

Based on joint works with Ohad Shamir, Michael I. Jordan, Tianyi Lin, Gilad Yehudai and Manolis Zampetakis.

Guy Kornowski

Guy Kornowski is a PhD student at the Weizmann Institute of Science, advised by Prof. Ohad Shamir. His research focuses on optimization and machine learning theory. He is a recipient of the Azrieli graduate fellowship.

Causal Data Integration

Prof. Brit Youngmann, Technion

Causal inference is fundamental to empirical scientific discoveries in natural and social sciences; however, in the process of conducting causal inference, data management problems can lead to false discoveries. Two such problems are (i) not having all attributes required for analysis, and (ii) misidentifying which attributes are to be included in the analysis. Analysts often only have access to partial data, and they critically rely on (often unavailable or incomplete) domain knowledge to identify attributes to include for analysis, which is often given in the form of a causal DAG. We argue that data management techniques can surmount both of these challenges. In this work, we introduce the Causal Data Integration (CDI) problem and discuss our proposed solution, which includes developing techniques to integrate input datasets with unobserved potential confounding variables, and causal DAG summarization.

Prof. Brit Youngmann

Brit Youngmann is an Assistant Professor at the Technion’s Faculty of Computer Science. Her research focuses on data management, causal reasoning, and responsible data management. She is developing automatized data tools to facilitate data analysis performed by scientists to accelerate scientific discoveries. Her research draws on ideas from data management and causal reasoning, making them practical for goal-oriented scientists working with real-life datasets. Before joining The Technnion, she was a postdoctoral researcher in the Data System Group at MIT, and received her Ph.D in Computer Science from Tel Aviv University.

Democratising Natural Language Processing: Overcoming Language and Domain Barriers in Low-Resource Environments

Dr. Yftah Ziser, University of Edinburgh

Natural language processing (NLP) has been revolutionised in recent years to the point where it is an inseparable part of our daily lives. The transition to transformer-based models allows us to train models on vast amounts of text efficiently, proving that scale plays a crucial role in improving performance. Unfortunately, many people worldwide are marginalised from getting access to high-quality NLP models, as the language they speak and the domains they are interested in count for only a tiny fraction of current state-of-the-art models’ training sets.

This talk will address the challenges, approaches, and opportunities for democratising NLP across different languages and domains by developing methods to improve NLP in low-resource scenarios. I will start by discussing how we can ease distribution mismatches to improve performance using representation learning. However, as NLP models become increasingly present in our lives, improving other crucial aspects beyond their performance, such as their fairness, factuality, and our ability to understand their underlying mechanisms, is essential. Therefore, I will also discuss using spectral methods to remove information from neural networks to reduce undesired attributes, such as bias, to increase fairness where sensitive data is scarce. Finally, I will explore future directions for making these models accessible toa broader audience by improving the aspects mentioned above in low-resource scenarios.

Dr. Yftah Ziser

Yftah Ziser is a Postdoctoral Researcher at the School of Informatics at Edinburgh University, hosted by Shay Cohen. He focuses on Deep-Learning methods for dealing with the resource bottleneck, which seriously challenges the worldwide accessibility of NLP technology. His research develops methods to improve low-resource models’ performance, fairness, and factuality while developing analysis methods for deepening our understanding of them. He co-organized the Domain Adaptation for NLP Workshop at EACL 2021.

Before joining the University of Edinburgh, Yftah worked as a research scientist at Amazon Alexa. Yftah obtained his PhD from the Technion, where he was advised by Roi Reichart.

Trustworthy and Informative Machine Learning for Accelerated Scientific Discovery

Dr. Mario Boley, Monash University in Melbourne, Australia

Machine learning promises to accelerate scientific theorydevelopment and discovery in a data-driven approach. However, to fulfil thispromise, methods have to a) provide an explicit human-readable form of themodelled relations and b) extrapolate well to unseen cases from only a fewexpensive data points. Modern deep learning systems, while producing impressiveresults in some areas, are fundamentally unsuited to meet these tworequirements, as they rely on vast quantities of parameters that interact in complicatedways and that need to be fitted using equally vast amounts of training data. Incontrast, additive models of simple basis functions can provide not only veryaccurate predictions for important scientific questions, they are also readilyunderstandable and testable akin to traditional empirical laws. I willdemonstrate this using three examples of my applied work in chemistry andmaterials science: modelling propagation rates in radical polymerisation,morphological outcomes of polymer-induced self-assemblies, and crystalstructure affinity of octet binary semi-conductors. Motivated by thesesuccesses, I will then discuss some recent methodological work on statisticaland algorithmic challenges in producing such trustworthy and informativemodels. In particular, I will show how a Bayesian treatment of linearregression leads to parameter estimates that are both statistically more robustand typically faster to compute than the usual cross-validation-based approach.Moreover, I show how a novel objective function and optimisation approach leadto a better accuracy/interpretability trade-off when iteratively assemblingadditive models within the commonly used framework of “gradient boosting”.

Dr. Mario Boley

Mario Boley is a Senior Lecturer and the Deputy Director ofResearch at the Department of Data Science and AI at the Faculty of IT ofMonash University in Melbourne, Australia. He is interested in trustworthymachine learning with a focus on efficient learning algorithms forinterpretable models and their application to accelerate scientific discovery,in particular in materials science and polymer chemistry. Mario obtained hisPhD in computer science in 2011 from the University of Bonn, Germany, for workin algorithmic order theory and the branch-and-bound algorithm. Subsequently,he held post-doctoral positions at the Fraunhofer Institute for IntelligentAnalysis and Information Systems, the Max Planck Institute for Informatics, andthe Fritz Haber Institute of the Max Planck Society for Materials Science. Hejoined the permanent academic staff of the Faculty of IT at Monash Universityin 2018.

The NVIDIA Deep Learning Institute (DLI) and the University of Haifa DSRC invites you to attend a free hands-on deep learning online workshops, exclusively for verifiable academic students, staff, and researchers.

Building Transformer-Based Natural Language Processing Applications 
20-21.12.2023 Wednesday-Thursday from 09:00 – 15:00

NVIDIA DLI offers hands-on training for developers, data scientists, and researchers looking to:

  • Learn how Transformers are used as the basic building blocks of modern LLMs for NLP applications
  • Learn how self-supervision improves upon the Transformer architecture in BERT, Megatron, and other LLM variants for superior NLP results
  • Leverage pretrained, modern LLM models to solve multiple NLP tasks such as text classification, NER, and question answering
  • Manage inference challenges and deploy refined models for live applications

About This Workshop – Building Transformer-Based Natural Language Processing Applications:




  • Python programming experience.
  • Basic understanding of neural networks, and a fundamental understanding of a deep learning framework such as TensorFlow or PyTorch.

The NVIDIA Deep Learning Institute (DLI) and the University of Haifa DSRC invites you to attend a free hands-on deep learning online workshops, exclusively for verifiable academic students, staff, and researchers.

Fundamentals of Deep Learning – 15.11.2023 Wednesday from 09:00 – 15:00

NVIDIA DLI offers hands-on training for developers, data scientists, and researchers looking to solve challenging problems with deep learning and accelerated computing.

An understanding of fundamental programming concepts in Python such as functions, loops, dictionaries, and arrays. Suggested materials to satisfy prerequisites: Codecademy Python course.

Applications of AI for Anomaly Detection – 16.11.2023 Thursday from 09:00 – 15:00

 NVIDIA DLI offers hands-on training for developers, data scientists, and researchers looking to solve challenging problems with deep learning and accelerated computing.


  • Professional data science experience using Python; experience training deep neural networks.
  • Experience training deep neural networks (DLI’s Getting Started with Deep Learning course is recommended)
  • Experience with data science with Python (Kaggle’s Intro to Machine Learning course is recommended)
Dr. David Deutsch, Dept. of Neurobiology
Prof. Daniel Sher, Dept. of Marine Biology
Dr. Uri Hertz, Dept. of cognitive sciences
Dr. Tomer sidi, Data Science Research Center
Dr. Iftach Amir, Data Science Research Center
Carmel campus, Madrega building, room 2007
4-13.9.2023, 08:30-18:30

Course page

Over the last few years data science tools, and especially machine learning (ML) techniques, have become increasingly popular and are now widely used in both the academia and the industry. The Haifa Data Science Summer School provides an overview of the variety of tools employed in machine learning. The course is aimed for students from diverse disciplines, with no prior experience using Machine-learning, and with a minimal experience in programming.
Our aim is to give each student enough theoretical and practical background to allow them to apply a wide range of ML techniques in their own work. Topics we will cover include exploratory data analysis, supervised learning (regression and classification, neural networks) and unsupervised learning (clustering, dimensionality reduction). For each topic, we start from an introduction, then discuss its uses, strengths, and limitations, then dedicate time for hands-on experience. Special attention will be devoted on
applying the methodologies introduced in class on empirical data using Jupyter notebooks coded in Python, with the machine-learning package ‘Scikit-learn’ (https://scikit-learn.org/ ).
It is an intense 8-day course (8:30am-6:30pm daily), with a final project. Students are expected to attend to all the meetings. Basic Python programming is a required. There will be a 4-hour prep meeting for students who wish to brush up on their Python skills, and we will also share some recommended online sources. Upon completion of the course requirements, registered students will be credited 3 academic points. The course will be given in English.
The course is open to 50 advanced (MSc and PhD) students, postdocs and research scientists from the university of Haifa and affiliated institutes (e.g. Oranim). Advanced undergraduate students are also welcome to apply and will be accepted based on free space. Participants credit are required to register and to submit a final course project taking the course for academic 

September12th, 2023, 09:00-19:30


The development of objective tools to identify and code facial and body movements in non-human species, as well as recent advances in Artificial Intelligence (AI) and Machine Learning, offer exciting new opportunities for the investigation of dog behavior, as well as for the development of novel applications related to dog welfare, wellbeing and health. This international workshop will provide opportunities for researchers and practitioners working on dog behavior, cognition and welfare to meet and exchange ideas.

Link to the workshop website


June 26th, 2023, 14:15-15:45

Zoom Meeting Recording

Title: AI-based Data Integration in the service of Oceanographic Research

Speaker: Dr. Tomer Sagi


The Oceanographic Data Integration Initiative (ODINI) is a joint project of
Daniel Sher, Tali Treibitz, and Yoav Lehahn of the University of Haifa with
Tomer Sagi of Aalborg University. The Data Integration component of
ODINI is comprised of several interesting research projects that will be
detailed in the lecture. The first is the main Ontology-based data
integration system, based on a novel schema matching engine utilizing the
textual descriptions of oceanographic datasets to improve the ability of a
schema matching system to identify the concepts in a dataset. The second
is the generalized entity resolution system, looking for relations between
datasets to avoid duplication of data and utilizing a stacked transformer
approach. The third is the central ontology of our system, for which
automated ontology construction and evaluation methods are being
researched. The fourth, is the extraction and reconstruction of datasets
from scientific data where new datasets are mined from scientific papers.


Dr. Tomer Sagi has been working with databases since he was 14 years
old. After several years as a consultant in the fields of Business
Process management and Business Intelligence, Dr. Sagi obtained a
Ph.D. from the Technion under the Supervision of Prof. Avi Gal. He
then proceeded to work as an industrial researcher at HP Labs Haifa
and subsequently at GE Healthcare. Between 2017 and 2022, Dr. Sagi
was a Lecturer at the University of Haifa, Department of Information
Systems. Since 2022, he is a Researcher and Adjunct Professor at
Aalborg University, the Department of Computer Science. Dr. Sagi’s
research centers on the interface between AI and Databases, with a
strong applicative focus, currently in three fields: Oceanography,
Middle-east History, and Medicine.

May 22th, 2023, 14:15-15:45

Zoom Meeting Recording

Title: AI4Biodiveristy project

Speaker: Prof. Ilan Shimshoni 


In this talk, we will describe the AI4Biodiveristy project we are undertaking as part of a Vatat supported grant for data science based research projects.


  • We envision a global hybrid human-computational system that provides timely information on a large number of species across the entire globe, allowing policy makers and conservation agencies to respond quickly with effective mitigation plans.
  • Furthermore, we foresee increasing people’s affinity to nature and thriving environmentally-conscious communities. As part of our vision, education systems would engage students with nature, shaping their awareness and attitude towards environmental issues.


  • To develop scalable tools and methods for biodiversity monitoring and estimation of populations of terrestrial vertebrate species, which would inform best practices in the field and will be adopted by nature protection agencies.
  • Furthermore, employing the biodiversity maps that we will produce, we aim to develop state-of-nature reports, which will inform policy recommendations and intervention plans for nature conservation.
  • Additionally, we plan working towards strengthening communities of, as well as collaborate with educators in developing curricular material related to conservation.


Prof. Ilan Shimshoni. Project’s PI. Department of Information Systems, University of Haifa

Ilan Shimshoni is a Professor of Information Systems. His research areas are computer vision, computer graphics and robotics. In recent years he has also been developing algorithms in CV for solving problems in Archeology, Rehabilitation, Geography, and especially ecology and agriculture. He served as the chair of the IS department and was an Associate Editor for IEEE TPAMI, the leading journal in the field of computer vision.


Title: A Social computing platform for Biodiversity monitoring

Speakers: Weaam Shaheen & Lior Koren 


In this talk, we aim to explore the utilization of human computation and machine learning techniques to expedite and enhance the tedious tasks faced by ecologists in monitoring biodiversity. By working towards the integration of human and machine intelligence, our intermediate goal is to create a crowd-computing platform accessible for citizen scientists, allowing them to augment or even substitute the efforts of a single expert, thereby expediting the process considerably.

Should our system prove successful, it will enable ecologists to harness algorithms and crowdsourced assistance to produce accurate and prompt assessments of nature’s status. Such data will be crucial for conservation organizations and authorities in formulating effective measures. Additionally, this approach is expected to strengthen the connection between citizen scientists and nature, while raising awareness about the importance of wildlife sustainability.


Keynote speakers: Manuel  Lima & Leticia Pozza


Artathon 2023

Tuesday 16:00-20:00

Wednesday 11:00-20:00

Thursday 11:00-20:00

Carmel Campus, Salah Building & via zoom (link will be supplied to registrants)



April 24th, 2023, 14:15-15:45

Zoom Meeting Recording

Title: Towards Automated Facial Landmark Detection for Animals

Speaker: George Martvel (martvelge@gmail.com)

Abstract: In this talk I will present the problem of facial landmarks detection on animals, a topic that has not received much attention in the field of computer vision, despite the mathematical interest of the problem itself and its numerous applications in the field of animal emotions recognition and well-being. I will provide an overview of the challenges in detecting landmarks on animals, and contrast the use of computer vision for humans and animals. Then, I will delve into the process of cascade detection of landmarks, using the example of cats to demonstrate how this technique can be applied to animals. Additionally, I will discuss the potential applications of this technology in fields such as animal behavior research and veterinary medicine.

Bio: I am a PhD candidate studying computer vision and animal-affective computing. The focus of my current research is to develop solutions for the detection of animal faces and facial landmarks for the classification of their internal state and emotions. I obtained my BSc and MSc in applied mathematics and physics from Moscow Institute of Physics and Technology and last year I started working on a PhD thesis at the University of Haifa.


Title: Using computational tools for revealing the neural basis of behavior in Drosophila

Speaker: Dr. David (Dudi) Deutsch  (ddeutsch1@univ.haifa.ac.il)


Understanding the neural basis of behavior is a major challenge in modern neuroscience. Drosophila melanogaster has long been served as an excellent model system for revealing the neural basis of behavior, due to the relative simplicity of this system, and the wide range of available genetic tools in this model. Recent advances in the field of machine learning are accelerating our understanding of the neural basis of social behavior in flies in two major directions. First, pose estimation is used for fine tracking of body parts of single or interacting flies, followed up by supervised and unsupervised algorithms for detailed and automated quantification of fly behaviors. Second, neural networks are being used for automatic detection of neurons in sub-micron resolution scans of the entire fly brain, therefore accelerating the process of building a fly ‘connectome’ – a description of all the connections between cells in adult or developing fly brains.

In my talk, I will first give some background on the use of machine learning based tools for studying the neural basis of social behaviors in flies. I will then describe two projects I completed during my post-doc, focusing on two aspects of social behaviors in flies: (1) sexual dimorphism in the circuits controlling social behaviors, and (2) the role of ‘internal’ brain states in controlling social decisions.

Last, I will describe some future directions of my recently established lab at the Neurobiology department, University of Haifa. Among them are deciphering the neural basis of social communication in complex, natural-like environments, and using the fly connectome for revealing principals regarding how the structure and connectivity of cells is related to their function.


David (Dudi) Deutsch received his B.Sc in electrical engineering from Tel-Aviv University in 2005. His M.Sc (direct path for outstanding students) was joint between Tel-Aviv university and the Weizmann Institute, focusing on electrostatic properties of adsorbed polar molecules. Following one year of travel in South America, he decided to move his scientific focus to understanding the mysteries of the brain. He joined the Neurobiology department at the Weizmann Institute of Science, where he did his Ph.D. under the supervision of Prof. Ehud Ahissar and Prof. Elad Schneidman. He studied how brains are actively controlling the flow of information that they collect from their environment (‘active sensing’), using the rat whisker system as a model.

Transitioning to his post-doctoral studies, he switched to studying the fruit fly Drosophila melanogaster, taking advantage of the tractability of the system, and the available genetic tools. In 2014 he joined the Murthy lab at the Princeton Neuroscience Institute. He studied social communication in males and females, focusing on two major questions: (1) What are the shared and sexually dimorphic circuits for the processing of courtship song in the male and female brains, and (2) How do internal motivational states modulate social behavior. Dudi opens his lab in the Neurobiology department at the University of Haifa in 2022, where he will study the neural basis of social communication.

April 19th, 2023, 11:00-13:00

Mr. Yaakov Diminsky
Calendar – Digital health startup
Eyal Enav

For more information and registration >>


March 20th, 2023, 10:00-14:00

Prof. Mor Peleg, Prof. Ilan Shimshoni, Prof. Einat Minkov, Dr. Itai Dattner, Dr. Dan Rosenbaum, Marcelo Feighelstein & Nir Lotan

For more information and registration >>

Understanding how metabolism works in cells, organisms or ecosystems is a highly interdisciplinary and complex task. Metabolic networks first need to be reconstructed, for example through automatic annotation of genome sequences. The resulting draft networks are curated, and can then be interrogated directly or used to interpret experimental evidence, for example transcriptomic, proteomic or metabolomics results. In this workshop, we will teach the fundamentals of genome sequencing and annotation, metabolic reconstruction and metabolic modeling, as well as approaches to analyze experimental results in light of cellular metabolism. The workshop includes 5 days of lectures and handson tutorials, using primarily the BioCyc platform (https://biocyc.org/). Some of the key topics we will discuss are:

  • How to design a denovo genome sequencing, assembly and annotation project
  • Metabolic network reconstruction and curation
  • Identifying metabolic pathways over-represented in experimental datasets (e.g. transcriptomics)
  • Construction and interpretation of genome-scale (constraint-based) metabolic models

The workshop is open to 50 advanced (MSc and PhD) students, postdocs and research scientists from around the world. Prerequisite courses include genetics, biochemistry, molecular biology, basic statistics and bioinformatics. Experience with R or other programming languages is useful but not a prerequisite. Registration is now open – please register online at https://forms.gle/TJcdPvQoXHUhwV6V7. Accepted participants will be notified during September. Participants taking the course for academic credit will then be required to register through the secretariat of their university and to submit a final course project.

Zoom Meeting Recording


Special Colloquium combining the DSRC seminar and the “Genomes to Metabolism” workshop


Title: Informatics as a Public Good

Speaker: Prof. Ida Sim

Digital innovations are transforming health care and research. Some of these
innovations originate from research funded by federal agencies and non-profit
organizations. Other innovations originate in the commercial sector and are
overwhelmingly built and scaled as for-profit companies. However, some
informatics technologies such as data, vocabulary, and interface standards are
better thought of as public goods, which commercial markets typically
undersupply. As ever more data are needed to drive artificial intelligence and
machine learning, informatics standards are becoming increasingly important for
breaking down silos and fostering more data sharing and integrated solutions.
This talk will describe the technical and policy landscape of health data sharing
in the United States, and will discuss how two non-profit organizations, Open
mHealth and Vivli, are advancing informatics-powered data sharing through
public goods approaches.

Ida Sim, MD, PhD is a primary care physician, informatics researcher, and entrepreneur. She is a Professor of Medicine and UCSF Director of the UCSF UC Berkeley Joint Program in Computational Precision Health. Her other UCSF positions include Director of Digital Health for the Division of General Internal Medicine and Co-Director, Informatics and Research Innovation at UCSF’s Clinical and Translational Sciences Institute. Dr. Sim is a global leader in the technology and policy of large-scale health data sharing.

Dr. Sim is a co-founder of Open mHealth, a non-profit organization that is breaking down barriers to mobile health app and data integration through an open software architecture. Open mHealth is an IEEE family of global standards. IEEE 1752.1 was officially approved in 2021. Dr. Sim has multiple grants from NIH, NSF, and AHRQ on mobile health methodology and digital health for primary care. In 2019, she co-developed CommonHealth, an open source software suite bringing to the Android ecosystem the equivalent of Apple Health’s ability to access and share EHR data.


Title: Expanding the Paradigm of Microbial Genome Annotation

Speakers: Dr. Peter Karp

The standard paradigm for computationally analyzing a newly sequenced microbial genome is to
assemble sequencer reads into longer contigs, to use gene-finding programs to identify the locations
of genes within those contigs, and to use sequence-similarity searches and HMM models to assign
gene functions. In recent years that paradigm has been extended in multiple respects through the
development of additional inference tools to extract additional information from the genome:
metabolic reconstruction techniques predict the qualitative metabolic network of the organism, and
generate a quantitative metabolic model for the organism. Pathway hole filling and reaction gap
filling identify genes coding for missing pathway enzymes, and identify missing metabolic reactions.
We also present tools for inferring transport reactions and protein complexes.

The preceding inference tools are available within the Pathway Tools software suite, and can be
applied to any newly sequenced genome. We have processed 20,000 genomes using these tools to
create the BioCyc collection of Pathway/Genome Databases. The BioCyc website provides an extensive
set of bioinformatics tools for searching and analyzing these databases, and leveraging them for
analysis of omics datasets. Genome-related tools include a genome browser, sequence searching and
alignment, and extraction of sequence regions. Pathway-related tools include pathway diagrams, a
tool for navigating zoomable organism-specific metabolic map diagrams, and a tool for searching for
metabolic routes that connect metabolites of interest. Regulation tools depict operons and regulatory
sites, as well as showing full organism regulatory networks. Comparative analysis tools enable
comparisons of genome organization, of orthologs, and of pathway complements. Omics data analysis
tools support enrichment analysis and painting of transcriptomics and metabolomics data onto
individual pathway diagrams and onto zoomable metabolic map diagrams. A new Omics Dashboard
tool enables interactive exploration of omics datasets through a hierarchy of cellular systems.

Peter D. Karp is the director of the Bioinformatics Research Group
within the Artificial Intelligence Center at SRI International.
Dr. Karp has authored 190 publications in bioinformatics and computer
science in areas including metabolic pathway bioinformatics,
computational genomics, scientific visualization, and biological
databases.  Karp developed the Pathway Tools software, the EcoCyc and
MetaCyc databases, and the BioCyc database collection.  He is a Fellow
of the American Association for the Advancement of Science and of the
International Society for Computational Biology.  He received the
Ph.D. degree in Computer Science from Stanford University in 1989, and
was a postdoctoral fellow at the NIH National Center for Biotechnology

Dear students and postdocs,

Doing a BSc, MSc, MA with thesis, PhD or Post-Doc means not only learning how to do scientific work, but also developing your career. How do I develop a scientific network and forge fruitful collaborations? How do I improve my communication with my advisors to fulfil my goals?

To address some of these questions, we developed a “soft skills” workshop, based on similar ones taught in leading institutes in Israel and abroad (and on our own experience).

For more information and registration please see here.

Keep safe,
Smadar, Daniel & the DSRC team

The Second International Israel Data Science Initiative Conference is taking off!
All the information and registration is here.
Please notice:
Early registration deadline – 4.11.2022.
Abstracts submission deadline – 30.09.2022.

Zoom Meeting Recording


Title: Contextualized ML-based Predictions for Clinical Settings 

Speaker: Kayla Schiffer (khs2138@cumc.columbia.edu‏)

Abstract: Many predictive models are developed and embedded into decision support tools, but fail to be  implemented or evaluated for safety, acceptance, and integration into clinical practice. We explore user-centered design for a predictive algorithm’s optimal integration into clinical settings. Our goal is to derive insights for an interventional clinical decision support (CDS) tool that is compatible with existing workflows and practices around evidence-based medicine in complex, team-oriented care environments to ensure safety and impact of a model’s predictive potential.
CDS systems’ lack of successful integration in clinical workflows has been attributed to the incompatibility of AI metrics with those for clinical pattern recognition; the absence of consideration for the sociotechnical systems in which they are embedded; the infancy of explainability methods and resulting misalignment with clinicians’ approach to care; among others. We develop a user-centered implementation of a risk score for delayed cerebral ischemia in patients with subarachnoid hemorrhage that avoids these pitfalls by focusing on contextualization of the prediction among existing clinical data. Our research sought to identify a presentation of the risk prediction as part of a contextualized CDS tool that relies on clinician pattern recognition to align with current practice. We aim to balance clinician and developer perceptions of an algorithm so as to not burden clinicians to dissect the ‘black box,’ and facilitate rapid assimilation of knowledge for effective decision making.

Bio: Kayla Schiffer grew up in New York. She obtained her BA at Barnard College, where she studied Medical Humanities and Computer Science. During her time at Barnard, her research focused on using patient generated data to support shared decision making with care providers for poorly understood chronic diseases. She then transitioned to industry, working at Veeva Analytics studying contemporary health care trends and their impact on consumer drug patterns. She is currently a second year PhD student in the Department of Biomedical Informatics at Columbia University, researching under Dr. Chunhua Weng. Her research focus is on knowledge discovery of temporal patterns of disease with an emphasis on interpretability in clinical settings. Kayla recently led a workshop at the American Medical Informatics Association annual conference focused on defining principles of social justice in health informatics.


Title: Show me your (people) and I’ll show you my (data): Psycholinguistic and computational approaches to studying bilingual language processing

Speakers: Prof. Shuly Wintner (shuly@cs.haifa.ac.il) and Prof. Anat Prior (aprior@edu.haifa.ac.il‏)

Abstract: Most individuals use more than one language in their daily lives, emphasizing the
importance of understanding bilingual language processing as a unique phenomenon,
distinct from monolingual language processing, which has been a major focus of
research up to date. One important facet of bilingual language processing are cross
language influences (CLI), namely when processing of one language is modulated by
knowledge of another language. In this talk, we will present three studies focusing on
how bilinguals process words that share form and meaning (cognates) across the two
languages. The first study used methods of corpus analyses and demonstrated that the
native language of bilinguals using English can be identified based on their use of
cognates with their native language (Rabinovich et al., 2018). The second study used
psycholinguistic lab methods and found robust CLI from Arabic and/or Hebrew when
bilinguals and trilinguals processed their non-native language, English (Elias et al., in
prep.; Schreiber, 2021). Finally, the third study is a joint effort using computational tools
to investigate a psycholinguistic question: Namely, are effects of CLI attenuated with
growing proficiency in the L2 (Native et al., under review)? We will discuss the added
value of using complementary research methods, and the benefits of collaboration.

Shuly Wintner is professor of computer science at the University of Haifa, Israel. His
research spans various areas of computational linguistics and natural language
processing, including formal grammars, morphology, syntax, language resources,
translation, and multilingualism. He served as the editor-in-chief of Springer's Research
on Language and Computation, a program co-chair of EACL-2006, and the general
chair of EACL-2014. He was among the founders, and twice (6 years) the chair, of ACL
SIG Semitic. He is currently the Chair of the EACL.

Anat Prior is an associate professor in the Faculty of Education at the University of
Haifa. I study individual differences in language learning, minority students’ literacy
performance, foreign language learning, interactions between first and second
language systems and domain general vs. specific bases of language processing. Our
goal is to better characterize the interactions between two languages (or more) in a
single cognitive system, and to identify the underlying mechanisms leading to
individual differences in language learning and processing. Ultimately, we aim for our
research to foster the development of effective instruction and intervention programs in
the domain of foreign language, and programs targeted at minority language students
in mainstream education.

Dear students and postdocs,

Doing a BSc, MSc, MA with thesis, PhD or Post-Doc means not only learning how to do scientific work, but also developing your career. How do I develop a scientific network and forge fruitful collaborations? How do I improve my communication with my advisors to fulfil my goals?

To address some of these questions, we developed a “soft skills” workshop, based on similar ones taught in leading institutes in Israel and abroad (and on our own experience).

For more information and registration please see here.

Smadar, Daniel & the DSRC team

Zoom Meeting Recording

Title: Photon ring autocorrelations

Speaker: Dr. Shahar Hadar (shaharhadar@sci.haifa.ac.il)

Abstract: Black holes strongly curve light rays, and therefore in their presence light sources connect to observers along multiple paths. As a result, brightness fluctuations in black hole observations must be correlated in intricate ways. In the talk I will describe a prediction for such correlations in black hole (time-dependent) images, which are expected to become available in the next few years thanks to the next-generation Event Horizon Telescope. I will first give general background on black holes.

Bio: I am a theoretical physicist studying gravitational physics, with a particular interest in black holes.  After completing my PhD at the Hebrew University under the supervision of Prof. Barak Kol, I held postdoc positions at Cambridge UK, the Max Planck Institute for Gravitational Physics, Potsdam, and Harvard University. Last year I joined the faculty of the department of mathematics, physics, and computer science of the University of Haifa at Oranim.


Title: Let there be Light: The first billion years of Cosmic History

Speaker: Prof. Saleem Zaroubi (saleem@astro.rug.nl)

Abstract: In this talk I will review the large effort to understand the first billion years of our cosmic history. During this period the first light emitting objects formed. These first objects ushered the Universe into a new era and changed the early pristine Universe to the galaxy, star and planet-filled one we observe around us. This effort is multi-disciplinary in nature with contributions from astronomers, engineers and computer scientists and involved peta bites of data and computer intensive calibration (essentially, an inversion problem).

Bio: Prof. Saleem Zaroubi grew up in Nazareth. He obtained his BSc and MSc from the Technion & his PhD from the Hebrew University, all in Physics. He was a Postdoctoral Fellow at the Physics and Astronomy departments, University of California at Berkley, and then a research associate at the Max Planck Institute from Astrophysics near Munich. He is a professor at the Kapteyn Astronomical Institute at the University of Groningen, The Netherlands, which he joined in 2004. Since 2016 he is also a professor at the Department of Natural and Life Sciences at the Open University. He is the chair of the newly established Astrophysics Research Center at the Open University (ARCO). In 2004 he founded, with two other Dutch colleagues, the LOFAR Epoch of Reionization project, one of the pioneering projects in the quest to explore the first billion years in the Universe’s history. He has just been awarded the 2022 Humboldt Research Prize for his research achievements by the Alexander von Humboldt foundation. Prof. Zaroubi writes about  science to the wider populic in Arabic and has recently published, also in Arabic, a book entitled “In the Beginning:  Physics, Philosophy and History of Cosmology”.

The NVIDIA Deep Learning Institute (DLI) and the University of Haifa DSRC invites you to attend a free hands-on deep learning workshop – Applications of AI for Anomaly Detection.
on 15.11.2022, Tuesday, from 09:00 – 15:00, exclusively for verifiable academic students, staff, and researchers.

NVIDIA DLI offers hands-on training for developers, data scientists, and researchers looking to solve challenging problems with deep learning and accelerated computing.

About This Workshop – Applications of AI for Anomaly Detection:


  • Professional data science experience using Python
  • Experience training deep neural networks

Full details and registration here.

All photos from the workshop here

The NVIDIA Deep Learning Institute (DLI) and the University of Haifa DSRC invites you to attend a free hands-on deep learning workshop – Fundamentals of Deep Learning.
on 8.11.2022 Tuesday from 09:00 – 15:00, exclusively for verifiable academic students, staff, and researchers.

NVIDIA DLI offers hands-on training for developers, data scientists, and researchers looking to solve challenging problems with deep learning and accelerated computing.

About This Workshop – Fundamentals of Deep Learning:

An understanding of fundamental programming concepts in Python 3, such as functions, loops, dictionaries, and arrays; familiarity with Pandas data structures; and an understanding of how to compute a regression line.

Full details and registration here.



All photos from the workshop here

Zoom Meeting Recording

The dynamics of animal social networks: Theory and practice

Speaker: Prof. Amiyaal Ilany

Abstract: The structure of animal societies impacts components of fitness, including survival, sexual selection, and reproductive success. Social structure is also instrumental to understanding of disease and cultural transmission, and the evolution of cooperation. Recent studies of animal sociality found multiple factors affecting social structure, but most of them used static social networks, ignoring temporal dynamics. I will describe a study of the environmental, individual, genetic, and structural factors that influence social network dynamics in spotted hyenas over 20 years, the largest such analysis to date. This study demonstrates that topological factors constrain the network dynamics, suggesting that some network structures are more stable than others. Next, I will present a general model of the dynamics of animal social networks, based on demographic stochasticity and social inheritance. This model can reconstruct the observed networks of multiple species. The model also suggests that heritability of social traits, and assortativity, the tendency to connect to similar individuals, can be explained as byproducts of social inheritance. Taken together, my results demonstrate the importance of clustering to the stability of social structure, and promote the move from observing patterns to understanding processes of animal sociality.

Short Bio:

Amiyaal Ilany grew up in the Arava Valley. He obtained a BSc in Computer Science & Biology, and MSc & PhD in Zoology, all at Tel Aviv University.

He was a Postdoctoral Fellow at NIMBioS (National Institute for Mathematical & Biological Synthesis; University of Tennessee, Knoxville), and then at the University of Pennsylvania.

Since 2016 he has been a faculty member at the Faculty of Life Sciences, Bar Ilan University, where he is now an Associate Professor.

In his free time he likes to watch rock hyraxes.

Dear all,
Please see this invitation from Tomer Sidi:

I’ll be hosting a private replay of an NVIDIA GTC session (GPU Technology Conference 19-22 September). I think you’ll find the below session particularly valuable. As we watch together, I’ll pause for questions and dialog with participants to address key points and share insights.

WPP41121c – Private Watch Party: A Deep Dive into RAPIDS for Accelerated Data Science and Data Engineering

Wednesday, 21 September, 2022


tomer watch party

RAPIDS enables data scientists and engineers to take advantage of GPU acceleration via simple, familiar Python interfaces.

In this session, We’ll dive into RAPIDS features, with an emphasis on newly released functionality in 2022, including SQL support, Triton and more.

Here are the steps to register:

  1. Click here to register for GTC if you haven’t done so yet.
  2. RSVP here for this watch party. You must RSVP yes to gain access as this is a private invitation only watch party.
  3. The watch party will appear in your schedule where you’ll be able to access the “Join Now Button” 15 minutes before the watch party starts.

I hope you’ll be able to join me and please feel free to invite your colleagues who might also be interested in attending the watch party session.

Best regards,

In July 2022 the DSRC is taking part of SICSS summer school:

The Summer Institutes in Computational Social Science

Bringing together graduate students, postdoctoral researchers, and junior faculty for 2 weeks of intensive study and interdisciplinary research.

July 3 to July 14, 2022.

All details of the program in the University of Haifa here.

More details about SICSS in general here.

We are delighted to announce an online seminar series as part of the AHRC funded research network Datasounds, datasets and datasense: Unboxing the hidden layers between musical data, knowledge and creativity. Starting January 2022, we will host guest speakers on the last Monday of the month on a range of topics relating to music, data and the gaps between them. The talks will take place in the afternoon (UK time). Abstracts, links for joining and specific time will be sent before each seminar. 
Feel free to email o.ben-tal@kingston.ac.uk if you wish to receive updates on the seminars or on the research network in general.

January 31 Renee Timmers (University of Sheffield) & Elaine Chew (IRCAM)

Renee Timmers’ current research projects investigate ensemble performance, in particular what visual and auditory nonverbal cues musicians use to coordinate and communicate with each other during performance.
Elaine Chew’s research centers on the mathematical and computational modeling of musical structures, with present focus on structures as they are communicated in performance and in ECG traces of cardiac arrhythmias.

February 28  Atau Tanaka (Goldsmiths University of London)

Atau Tanaka conducts research in embodied musical interaction. This work takes place at the intersection of human computer interaction and gestural computer music performance. He studies our encounters with sound, be they in music or in the everyday, as a form of phenomenological experience. This includes the use of physiological sensing technologies, notably muscle tension in the electromyogram signal, and machine learning analysis of this complex, organic data.

March 28 Blair Kaneshiro (Stanford University)

Blair Kaneshiro’s research focuses on using brain and behavioral responses to better understand how we perceive and engage with music, sound, and images. Other research interests include music information retrieval and interactions with music services; development and application of novel EEG analysis techniques; and promotion of reproducible and cross-disciplinary research through open-source software and datasets.

April 25  Anna Xambo (De Montfort University)

Anna Xambo envisions pushing the boundaries of technology, design, and experience towards more collaborative, egalitarian and sustainable spaces, what I term intelligent computer-supported collaborative music everywhere. My mission is to do interdisciplinary research that embraces techniques and research methods from engineering, social sciences, and the arts for creating a new generation of interactive music systems for music performance and social interaction in alignment with Computer-Supported Collaborative Work (CSCW) principles. 

May 30 Jeremy Morris (University of Wisconsin-Madison)

My research focuses on new media use in everyday life, specifically on the digitization of cultural goods (music, software, books, movies, etc.) and how these are then turned into commodified and sellable objects in various digital formats. My book, Selling Digital Music, Formatting Culture, focuses on the shared fate of the computing and music industries over the last two decades and my recent co-edited collections examine Apps (Appified, 2018) and Podcasting (Saving New Sounds, 2021).

June 27 Psyche Loui (Northeastern University) 

Psyche Loui’s research aims to understand the networks of brain structure and function that enable musical processes: auditory and multisensory perception, learning and memory of sound structure, sound production, and the human aesthetic and emotional response to sensory stimuli. Tools for this research include electrophysiology, structural and functional neuroimaging, noninvasive brain stimulation, and psychophysical and cognitive experiments

The Datasounds, datasets and datasense: Unboxing the hidden layers between musical data, knowledge and creativity network aims to identify core questions that will drive forward the next phase in data-rich music research, focused in particular on creative music making. The increased availability of digital music data combined with new data science techniques are already opening new possibilities for making, studying and engaging with music. This direction is only likely to speed up upending many current practices, opening up creative avenues and offering new opportunities for research. However, the rapid technological progress with new techniques producing surprising results in rapid succession, is often disconnected from the knowledge and knowhow gained by musicians through creativity, practice and research. By bringing together researchers and practitioners from different underlying disciplines and with a wide range of expertise the network will enable a better foundation for future research. Performers, composers, and improvisers will contribute through embodied knowledge and practice-based methods; researchers in psychology will bring insights about cognitive, affective and behavioural processes underpinning musical experience; and data scientists will add analytical expertise as well as relevant theories, methods and techniques.  These will lead to significant conceptual breakthroughs in data driven approaches and technologies applied to music.
The network is lead by Oded Ben-Tal (Kingston University) in partnership with Federico Reuben (York University), Emily Howard (PRiSM, Royal Northern College of Music),  Robin Laney (Open University), Nicola Dibben (University of Sheffield), Bob Sturm (Royal Institute of Technology, KTH, Sweden) and  Elaine Chew (IRCAM)
For registration and more details please contact Oded in o.ben-tal@kingston.ac.uk.

On Wednesday June 15, 2022, we hosted a meeting with Helmholtz delegation from Germany.
During their visit we presented our research center and learned about theirs, had an interesting discussion, and 3 researchers from our center presented their work.

We hope it will be the start of a fruitful collaboration with the centers of Helmholtz.


14:30 – Prof. Mor Peleg

Title: Introduction to the Data Science Research Center (DSRC). Prof. Mor Peleg, Dept. of Information Systems and Head of the DSRC

Short Bio:
Mor Peleg holds a BSc and MSc degrees in biology and PhD (1999) in Information Systems, from the Technion and post-doctoral studies (1999-2003) at Stanford’s Medical Informatics program. She is Full Professor at the Department of Information Systems, University of Haifa, which she joined in 2003, Head of the University of Haifa’s Data Science Center, former Chair of the Department of Information Systems, and founder and past Chair of the BSc Data Science program. She is Editor-in-Chief of the Journal of iomedical Informatics, and is Fellow of the American College of Medical Informatics and of the International Academy of Health Sciences Informatics. Prof. Peleg is internationally renowned in the area of clinical decision support, with a focus on using ontologies to integrate patient data with clinical knowledge. She led the large-scale European project MobiGuide, providing sensor-monitoring and evidence-based personalized decision-support to patients any time everywhere. Her current research exploits AI and Big Data for a new model of cancer care.

14:45 – 
Dr. Itzik Klein

Title: Data-Driven Autonomous Navigation and Sensor Fusion

The purpose of navigation is to determine the position, velocity, and orientation of manned and autonomous platforms, humans, and animals. Obtaining accurate navigation commonly requires fusion between several sensors, such as inertial sensors and global navigation satellite systems, in a model-based nonlinear estimation framework. Recently, data-driven approaches applied in various fields show state-of-the-art performance, compared to rule- or model-based methods. This talk will address data-driven based navigation algorithms, derived at the autonomous navigation and sensor fusion lab that enhance common navigation and estimation tasks. The algorithms include: 1) autonomous underwater vehicle navigation 2) indoor positioning using the smartphone’s inertial sensors 3) adaptive nonlinear filtering with model uncertainty for autonomous vehicles 4) quadrotor dead reckoning 5) data-driven denoising of accelerometer signals.

Short Bio:
Itzik Klein received the B.Sc. and M.Sc. degrees in Aerospace Engineering from the Technion – Israel Institute of Technology, Haifa, Israel, in 2004 and 2007, respectively, and a Ph.D. degree in Geo-information Engineering from the Technion – Israel Institute of Technology, in 2011. He is currently a Senior Lecturer (Assistant Professor), heading the Autonomous Navigation and Sensor Fusion Lab, at the Hatter Department of Marine Technologies, University of Haifa. Itzik serves as an IEEE Senior Member and has exceptional experience in navigation systems and sensor fusion. He has worked in the industry on navigation related subjects for more than 15 years at leading companies in Israel, prior joining University of Haifa. His research interests include data-driven based navigation, novel inertial navigation architectures, autonomous underwater vehicles, sensor fusion, and estimation theory.

14:55 – Prof. Doron Chelouche

Title: How Empty is Space? The Case for Astrostatistics and Astroinformatics

Astronomy is the oldest discipline of the natural sciences. With the advance in observing capabilities of ground and space telescopes, astrostatistics and astroinformatics are emerging fields of science. These have the potential to shed light on fundamental scientific questions pertaining to the history of the universe and its constituents, the birth of death of stars, and the properties of planets outside our solar system. In this talk some of the outstanding questions in the field will be reviewed with a particular emphasis on the ability of astrostatistics to detect cosmic dust at the edge of the universe with implications for the emergence of life in the universe.

Short Bio:
Doron Chelouche is a professor of physics and vice-dean of research at the faculty of natural sciences. He holds a Ph.D. from Tel-Aviv University (2005). He was the first Israeli to have been awarded a Chandra postdoctoral fellowship, and became a member of the school of natural sciences at the Institute for Advanced Study in Princeton (2005-2008). He later joined the Canadian Institute for Theoretical Astrophysics at the University of Toronto before assuming a position at the University of Haifa in 2009. His research interests are in the fields of astronomy and astrophysics, focusing mainly on the study of supermassive black holes and of diffuse material across space. His research has been supported by multiple grants from the EU (Marie Curie IRG), DFG, Israeli Science Foundation, Planning and Budgeting Committee, and NASA.

15:05 – Dr. Anna Brook

Title: Machine Learning for the Geosciences

Advances in sensor technology reveille new opportunities for identifying and predicting changes in Earth’s ecosystems in response to climate change to understanding interactions among the ocean, atmosphere, and land and provided alternatives to the conventional (broadband, low to medium resolution multispectral) satellite sensors. Hyperspectral airborne sensors e.g. AVIRIS, together with active sensors such as light detection and ranging (LIDAR), Radio Detection and Ranging (RADAR), and Terrestrial Laser Scanning (TLS) have shown benefits for uniquely challenging problems in Geoscience. The developments in Unmanned Aerial System (UAS) technology introduce advances in sensor performance and miniaturization. These systems are capable of carrying multispectral, high spatial resolution spectrometers, LiDAR, and thermal sensors. UAS products have the potential to determine initial and extended fire impacts and offer land managers options for high spatial and temporal pattern recognition and anomaly detection.

Consequently, these big data (multi-sensor spectral data) can no longer be analyzed by any of the conventional methods. Geosciences need to process large and rapidly increasing amounts of data to provide more accurate, less uncertain, and physically consistent inferences in the form of prediction, modeling, and understanding the complex Earth system. To get the most out of the rapid growth and diversity of Earth system data, we face two major tasks in the coming years: extracting knowledge from the data, and deriving models that learn much more from data than traditional data assimilation approaches. Advanced machine learning (ML) proposes methods that optimize their performance iteratively by learning from the data. These approaches demand detailed information, thus spectral reduction is not necessary and even undesirable. Such methods can be predictive or distinguish between different classes or patterns. Among the main subsets of ML, applications of genetic and biological programming in the geoscience and remote sensing domain are very new and restricted to a few areas.

Multi-source, multi-scale, high-dimensional, complex Spatio-temporal relations, including non-trivial and long-distance relationships between variables. The increasing demand for probability models, which are more advanced classifiers, has been met with alternatives: Artificial Neural Networks (ANNs) 5, Support Vector Machines 6, Wavelet Analysis, Logistic Regression, and decision trees, as classifiers for high dimensional data. ML techniques and recently established deep learning (DL) models have been proposed for spatial, spectral, and temporal imaging data classification. Deep learning is well-positioned to address the multi-source, multi-scale, high-dimensional, complex data challenges, and network architectures, and developed algorithms produce approaches that address both spatial-spectral and temporal context at different scales.

Convolutional neural networks (CNN), at the front of the current state-of-the-art in DL, first achieved successes in the field of image recognition and classification. The retrieval algorithms are actively underway in the field of supervised DL. ML in general and DL in particular offer promising tools to build new data-driven models for components of the Earth system and thus to build our understanding of Earth. Data-driven ML approaches to geoscientific research will not replace physical modeling but strongly complement and enrich it. Thus hybrid modeling approach, coupling physical process models with the versatility of data-driven ML should be our next step.

Short Bio:
Anna Brook is the head of the Spectroscopy and Remote Sensing Laboratory at the department of Geography and Environmental Studies, University of Haifa. She received a Ph.D. in Environmental Sciences with a thesis entitled “Reflectance Spectroscopy as a Tool to Assess the Quality of Concrete in situ” from the Porter School of Environmental Studies at Tel-Aviv University in 2011. Her research focus is on questions which drive technological, environmental and social change. Her aim is to bridge the gap between machine learning and geoscience for sustainability and environmental management at the national and international scales Her primary scientific interest is in developing hybrid approaches by coupling physical processes with the versatility of data-driven machine learning to better understand the ecosystems, biodiversity, dynamic processes and environmental responses to stressors, emphasizing sustainability and decision support system development tools.


15:15 – Helmholtz’s Presentation

IsraHCI – The Israeli conference on research in Human Computer Interaction

June 15th, 2022. 09:00-17:00

IsraHCI was a one-day conference held at the Hecht auditorium at the University of Haifa (see https://www.israhci.org).
The conference brings together academics and practitioners who are interested in the broad field of human-computer interaction to talk and hear about the latest research in the area.

The conference included a keynote speaker, a panel, 13 talks, 16 posters and 7 demos. 190 people registered to the conference, coming mostly from Israel, with several guests from the U.S.

The keynote, given by Prof. David Karger from MIT, discussed systems and methods to empower participants of online discussion forums, providing them with richer, more expressive tools. The panel which included Prof. David Karger, Prof. Tsvi Kuflik, Dr Aya Soffer (from IBM) and Dr. Noa Mor was about responsible and explainable AI, discussing how to design AI systems that would be more ethical and understandable to users.

The conference was organized by Prof. Joel Lanir from the information systems department and Dr. Sarit Szapiro from the education faculty.

Zoom Meeting Recording

Data Science and Health

Speaker: Dr. Hadar Fisher

Title: Does Feeling Bad, Lead to Feeling Good? Patterns of change in emotions during treatment for depression

Abstract: Depression is one of the most prevalent mental health disorders and the leading cause of disability worldwide. One of the leading theoretical models to understand the mechanisms underlying symptoms of depression suggests that the emotional impairment observed in depression results from emotional avoidance. From this perspective, to be modified, negative emotions should first be expressed. Most psychological treatment modalities incorporate interventions aimed at reducing emotional avoidance and increasing the experience and expression of emotions in adaptive ways. However, scant attention has been paid to how emotions are modified during psychological treatment.
In this talk, I will present a study in which we used automatic facial recognition tools and advanced statistical methods to explore whether targeting emotional avoidance during psychological treatment can change the way emotions are experienced and expressed by patients diagnosed with depression.

Short Bio: Dr. Hadar Fisher is a licensed clinical psychologist and a postdoctoral fellow at the Department of Statistics and the Department of Psychology at the University of Haifa. She is researching the role of experiencing and expressing emotions as a mechanism of change in psychological treatment.


Speaker: Dr. Yael Garten

 How personal mobile devices and machine learning can provide early warning signals to potential health issues

Abstract: Large-scale healthcare innovation is happening right on the devices we use every day. Revolutionary sensors in iPhone and Apple Watch can provide health and fitness metrics. Intuitive apps help users understand changes in their health while protecting user data. Virtual, large-scale medical studies bring together academic researchers, medical institutions, and healthcare organizations to accelerate innovation. Leveraging these tools and advancements in machine learning, we’re creating opportunities to detect potential health issues early on and empower users to better manage their health.

Short Bio: Dr. Yael Garten is Director of AI/ML Data Science & Data Engineering at Apple where she leads an organization of engineers, data scientists, data visualization experts, and product managers. Her team partners with Apple’s hardware, software and services teams to measure, understand, and optimize the user experience across Apple products and services, creating powerful data-informed machine learning based product experiences. Yael is a technology business leader with 20 years of experience in data science, machine learning, consumer product strategy, natural language processing and biomedical informatics. She serves on the Board of Directors of Levi Strauss & Co.

Previously, Yael was Director of Data Science at LinkedIn. Prior to that she was a research scientist at Stanford University; her research focused on information extraction and semantic understanding of pharmacogenomic text, to enable clinicians to make better decisions. She champions data culture & data quality with a focus on privacy & ethics, and has developed organizational best practices and tools to democratize data within companies. Yael has a Biomedical Informatics PhD from Stanford University School of Medicine, an MSc from the Weizmann Institute of Science in Israel, and a BSc in Computational Biology from Bar Ilan University. She lives in California with her husband and 3 children.

Digital Humanities Hackaton 2022

Award-winning competition in the field of preservation, accessibility and digital research of material, artistic, and cultural heritage treasures.

Dates: 19-20.5.22

Location: The lobby of Younes & Soraya Nazarian library


All the information (in hebrew) and registration is here.

AI research centers from Lower Saxony, Germany – ceremony for signing a Memorandum of Understanding

On Sunday May 1, 2022, a delegation including about 40 participants, including the Minister of Research of Lower Saxony, Björn Thümler, the state secretary in his ministry, Sabine Johannsen, several presidents / vice presidents of Lower Saxony universities, and heads of Artificial Intelligence (AI) research centers – Prof. Dr. Wolfgang Nejdl, Prof. Dr. med. Dr.-Ing. Michael Marschollek, founders of the Leibniz AI Lab, and others.

During that visit, a Memorandum of Understanding (MoU) between the Lower Saxony AI research centers and the University of Haifa’s Data Science Research Center will be signed by the Minister of Research and our University’s President, Prof. Ron Robin.

The collaboration will allow to promote excellence-oriented international research cooperation more strongly, to increase the transfer of knowledge and technology, and to establish sustainable international knowledge and innovation networks.

Schedule lower sax Photos from the event – here.

Zoom Meeting Recording

Big Data and Communication

Speaker: Dr. Chih-Chi Lee (DSRC Postdoctoral Fellow, University of Haifa)

Title: Host-microbe interactions and the social microbiome in ants

Abstract: Microbes (bacteria, archaea, viruses, and microbial eukaryotes) inhabit many parts of an animal body. For example, on the skin or exoskeleton, in guts, or even within a cell. These microbes sometimes provide vital functions essential for host survival or influence host behaviors. How a microbes’ community (or microbiome) is maintained and what its impact on a group-living animal is an open question. Ants are group-living social insects whose colony size could be determined by genetic mechanisms, therefore suitable to untangle intricate microbe-host interactions in various group sizes. In the first part of this talk, I will introduce how we used bioinformatic analysis on genomic data to uncover a tightly-link relationship between intracellular bacteria and its host ant across Southeast and East Asia. Next, I extend my research on microbial metacommunity in ants between different social behaviors to understand their relationships. I will discuss future research on microbiome-host interactions in the last part of the talk.


Speakers: Prof. Daphne Raban
(Associate professor in the School of Business Administration and Academic Head of the Library, University of Haifa)
Lina Portnoy
(Doctoral candidate at the School of Business Administration, University of Haifa)

 Self-Presentation in Online Marketplaces

Abstract: Prior research has shown that social interaction is important for information exchange online. Given that social interaction occurs between two people, this seminar will focus on the contribution of the personal dimension to the exchange through the lens of self-presentation. Self-presentation includes intentional and unintentional expressions that influence others’ perception of us, offline and online. We will present several studies showing under what conditions self-presentation is beneficial for direct economic rewards, what are some quantitative measures and limits of self-presentation, and how it contributes to social capital and for generating business leads. The data used in the studies was scraped from a Q&A website, a corporate social network and from Etsy.

Zoom Meeting Recording

Data Security and Mobility

Speaker: Shaul Kfir (Digital Asset)

Title: Cross-organization coordination – the use of blockchain and associated technologies to synchronize across organizations while maintaining data privacy

Abstract: The popularity of cryptocurrencies sparked the massive commercialization of multiple technologies that promise to help coordinate data and workflows across organizational boundaries while maintaining strong data privacy and operational controls. In this talk, we’ll cover some of these technologies, what problems they solve for companies and governments, and where they are in the adoption lifecycle.

Shaul Kfir (Digital Asset)

Short Bio: Shaul is one of the Founders of Digital Asset. He is a software engineer with a background in research cryptography, implementing Zero-Knowledge Proof protocols for Secure Computational Integrity and Privacy. Previously, he served as CTO for two startups in Tel Aviv. Shaul is a former Lieutenant Commander in the Israeli Navy. He was also a team member of the SCIPR cryptography lab and a visiting scientist at MIT CSAIL


Speaker: Dr. Aneta Lisowska, senior postdoctoral researcher at Sano Center for Computational Medicine and Poznań University of Technology

Title: Personalizing Digital Health Interventions with AI

Abstract: Digital health interventions (i.e., interventions implemented through digital technologies such as smartphones, websites, emails, wearable devices) have been shown to improve outcomes in patients suffering from chronic diseases, allow remote access to effective treatments and support changing health risk behaviors such as inactivity, unhealthy diet or substance abuse.

Machine learning (ML) is often used to personalize interventions. To facilitate intervention compliance and maximize its effectiveness, it is important to provide the right support at the right time.

In this talk, I will present recent work on designing and personalizing digital health interventions with case studied from the Horizon 2020 Cancer Patient Better Life Experience project (CAPABLE).  In particular, we rely on Fogg’s behavior theory that states that habits are formed when people are motivated, have the capability of performing the activity and there is a Trigger for the habit. The activities we explore are from the domains of mindfulness, physical activity and positive psychology. We explore different machine learning approaches to find the best time to intervene (fitting the Trigger component). We will also examine the factors affecting the effectiveness of the intervention based on the actual outcome of the real-world pilot study. Finally, I will discuss future research directions.

 Short Bio: Dr. Aneta Lisowska is a senior postdoctoral researcher at Sano Center for Computational Medicine and Poznań University of Technology, developing AI methods for digital health interventions.

Aneta’s research journey began in Scotland where she obtained a BSc in Computing and Cognitive Science from the University of Dundee and completed an industrial doctorate at Toshiba Medical Visualization System Europe and Herriot-Watt University. After obtaining an Engineering Doctorate, she joined the AI team at Canon Medical Research Europe where she worked on AI methods for text and medical image analysis, delivering numerous patents utilizing deep learning methods in healthcare applications, and leading research projects focused on Active and Continual Learning.

After over 6 years of working in the medical industry on AI clinical diagnostic tools, Aneta decided to change the direction of her research towards designing AI-based applications for ubiquitous mobile devices.  At the start of 2021 she joined Poznań University of Technology where she develops machine learning models for personalised coaching and decision support. In particular, she works on ML-based approaches supporting patients with treatment adherence and development of positive health habits.

In late 2021, Aneta also joined Sano, where she works on chronic illness prevention applications.

Personal Website

University of Haifa Data Science Research Center invites you all to a unique event:

Data Science for Society
 | מדעי הנתונים למען האנושות

10-12 March 2022
in the cafe-gallery Kibutz Galuyot 91, Haifa

Come to see exhibits which use different art forms to describe researches
that combine data science methods with subjects and fields that help society:
mental health and ancient cultures.

Come to hear all the information and explanations from the researchers and students
themselves who work on those fascinating projects.

On Thursday, 10.3.22@19:00 we’ll start with a festive launch evening!
The evening will include live performance of “The Scrollers”,
happy bar and delicious buffet, all free for the guest of the exhibition
+ we’ll give vouchers for free Coffee and Pastry throughout the exhibition days.

The exhibition will also be open at:
Friday 11.3.22, between 9:00-15:00
Saturday 12.3.22, between 9:00-15:00

Zoom Meeting Recording

Data science and Behavioral Science

 Dr. Liron Rozenkrantz, Department of Psychology and DSRC Postdoctoral Fellow, University of Haifa

Title: Linking Data-Science and Clinical Psychology to Optimize the Psychotherapeutic Process 

Abstract: In psychotherapy, patients’ expectations of therapeutic improvement (“expectancy”) have a profound effect on treatment outcomes, with higher initial expectations generally leading to better treatment outcomes. Thus, elucidating the predictive power of patients’ expectancy provides a means for optimizing treatment efficacy. However, the mechanisms by which expectancies exert their effect are poorly understood – as well as their dynamic nature over the course of treatment. Moreover, conventional statistical methods are not sufficient to fully explore these dynamics.

In this talk, I will first present the goal of this project: to apply advanced data science tools on a uniquely rich and dynamic dataset of a 16-session psychotherapy treatment in depressed patients, in order to understand how expectancy changes the course of treatment both behaviorally and physiologically – and for whom. Then, I will present the steps we took to elucidate the unexplored dynamics of expectancy’s effect on treatment outcome, as the grounds to study its mechanisms. Finally, I wish to discuss the next steps we plan to take, as well as the challenges and the opportunities in this link between data science and clinical psychology.


Speaker: Dr. Nehama Lewis-Persky, Department of Communication, University of Haifa

Title: Merging data science with behavioral science to predict and influence individual health behavior at scale: Opportunity, Obstacles and Ethics 

Abstract: The talk outlines how theories of behavior change, information processing, and persuasion can be used together with AI to influence individual health behaviors at scale. Behavioral science offers validated concepts and models that can be used to understand and influence attitudes and behavior, but is often tested in narrow academic contexts. In contrast, data-driven models can be used to predict behavior with increasing accuracy, but do not tell us why people decide to perform a behavior, or the decision factors that might influence this. By combining both approaches, it is possible to understand, predict, and influence individual behavior at scale. However, this approach also raises critical questions and concerns about the ethics of influence, particularly in the context of health.

Join us to the Annual Symposium of MLIS– the Technion AI center and TCE Center, on February 24, 2022, at ELMA Arts Complex [Zichron Ya’acov]. AI is now THE Buzz word, but what are the true current capabilities? What is state-of-the-art and what can be implemented? In this symposium, we differentiate fact from (current) fiction through a series of lectures and thematic workshops presented by Technion researchers. Conference website

IDSI in collaboration with Responsible AI, Law, Ethics & Society, hold a workshop in the matter of Teaching & Learning Responsible AI. The workshop will held in Monday, January 10, 2022, 9:00-17:00, in Tel Aviv University. All details in IDSI website.

The Israel Data Science Initiative is happy to announce its first international conference to mark the consolidation of the Data Science Research Centers that were established in 7 of Israel’s research universities.
Through these centers, and under the coordination of the IDSI, researchers from core and satellite Data Science disciplines have begun to have an impact on research and education in Israel, on collaboration with industry, and on forging international ties.

The conference is open to presentations and posters from Israeli and international researchers and practitioners.
Presentation and poster topics may include contributions from core DS fields such as AI, NLP, Statistics, ML, Deep Learning and from their applications to research in the biological, physical and social sciences as well as in the humanities.

For more details and to register for Early Bird, please visit our conference website on the IDSI website, here.

Zoom Meeting Recording

Data Science and Sustainability

Speaker: Lior Greenspoony, The Department of Plant and Environmental Sciences, Weitzman Institute.

Title: The Global Biomass of Wild Mammals

Lior Greenspoon*1, Eyal Krieger*1, Yuval Rosenberg1, Yinon M. Bar-On1, Uri Moran1, Tomer Antman1, Shai Meiri2, Uri Roll3, Elad Noor1 and Ron Milo1

1Department of Plant and Environmental Sciences, Weizmann Institute of Science, Rehovot, Israel

School of Zoology & Steinhardt Museum of Natural History, Tel Aviv University, Israel

Department of Desert Ecology, Ben Gurion University of the Negev, Israel

Mammals are icons of conservation efforts, yet there is no rigorous estimate available for their global biomass. Biomass as a metric allows us to compare species with very different body sizes, and can serve as an indicator of wild mammal presence on a global scale. We compiled estimates of the total abundance of several hundred mammal species (i.e., the number of individuals) from the available data, and used these estimates to build a model that infers the total biomass of terrestrial mammal species for which the global abundance is unknown. Here, we present a thorough assessment, arriving at a total wet biomass of ≈26 million metric tons (≈26 Mt) for all terrestrial wild mammals, the equivalent of about 3 kg per person on earth. The primary contributor to the biomass of wild land mammals is the white-tailed deer, followed by the wild boar, African elephant and eastern grey kangaroo. We find that even-hoofed mammals (such as deer and boars) represent ≈45% of the combined mass of terrestrial wild mammals. In addition, we estimate the total biomass of wild marine mammals at ≈47 Mt, with baleen whales comprising ≈60% of the mass of marine mammals. In order to put wild mammal biomass into perspective, we additionally estimate the remaining members of the class Mammalia. The total mammal biomass is dominated by livestock (≈630 Mt) and humans (≈390 Mt). This work is a provisional census of wild mammal biomass on Earth, and can serve as a benchmark for human impacts.

 Dr. Shlomit Sharoni, postdoc at the University of Haifa.

Title: Elucidating phytoplankton physiological performance in the ocean

Phytoplankton are unicellular, drifting microorganisms that form the base of the marine food web. Their activity contributes ~50% of the global annual net primary productivity. Phytoplankton growth depends on nutrient availability. However, nutrients in the marine environment are often scarce, particularly in the open ocean. Thus, a major open question is whether phytoplankton communities in the ocean mostly experience nutrient stress or are well adapted to prevailing conditions. These contrasting physiological states affect the elemental composition of phytoplankton cells. To differentiate between these two alternate scenarios, we thus used observational datasets on phytoplankton elemental ratios and a model of phytoplankton physiology. We find that on average, phytoplankton are well adapted to their environment, even in nutrient poor regions, likely due of recurrent selection of the most fitted genotype. This mechanism involving shifts in phytoplankton communities likely sustains an effective flow of matter and energy through biogeochemical pathways.


Lior Greenspoony, The Department of Plant and Environmental Sciences, Weitzman Institute.

Lior Greenspoon is a PhD student at Ron Milos’ lab the Weitzman Institute, at the Department of Plant and Environmental Sciences. Ron’s lab use a combination of computational and experimental synthetic biology tools to study the central carbon metabolism in quantitative terms.
In her undergraduate studies Lior majored in CS and Geology. She is using DS methods to study ecological questions.


Dr. Shlomit Sharoni, postdoc at the University of Haifa.

Dr. Shlomit Sharoni, postdoc at the University of Haifa, working with Yoav Lahan from Marine Biology and Tali Treibitz from the Dept. of Marine Technologies. She uses DS methods to study marine biology ecological questions. Shlomit holds a BSc in Chemistry from the Hebrew University. Her MSc and PhD are from the Weitzman Institute at the Department of Earth and Planetary Sciences, where she worked on Modeling Global Biogeochemical Cycles, Marine Biology and Ecology.

Dear all, MLIS, the Technion AI center, in collaboration with TCE – would like to invite you to participate in the annual TCE-MLIS conference which will initiate a cross-sector discussion on best-practices for the implementation of AI technologies in Israeli society. The tentative program includes 2 panels and 4 workshops:
  • Panel #1 | AI implementation in the governmental and public sector.
  • Panel #2 | AI implementation in the business sector.
  • Workshop #1 | Obstacles and opportunities for AI in education.
  • Workshop #2 | The role of AI in regional and municipal government.
  • Workshop #3 | Cooperation models between Academia and Industry in the AI age.
  • Workshop #4 | Data flow and regulation between data players.
  We are very excited about this in-person meeting and looking forward to seeing you! Details on registration and schedule are provided below and in the PDF file. Do not miss the EARLY BIRD reg. by 15.11.2021 All details in IDSI website.   image003

מספר הולך וגדל של החלטות בנוגע לחיי היומיום של בני אדם נשלטות על ידי אלגוריתמי בינה מלאכותית בתחומים הנעים משירותי בריאות, תחבורה וחינוך ועד קבלה למכללות, גיוס עובדים, מתן הלוואות ועוד תחומים רבים נוספים. למרות שלהתקדמות טכנולוגית זו יתרונות רבים בשיפור חיי היומיום, מחקרים עדכניים הראו שקבלת החלטות אלגוריתמית עשויה להיות נוטה מטבעה לחוסר הוגנות ואפליה. 

הצטרפו לכנס מרתק בו מיטב המומחים בתחום יסקרו את הקשיים בהטמעת הוגנות ואחריות בבינה מלאכותית וכן ידונו בפתרונות אפשריים

לפרטים נוספים ורישום מהיר >>  לחצו כאן

**כנס וירטואלי ללא עלות | 9.12.2021 | 10:30-15:30**



Zoom Meeting Recording

DSRC Hackathons

 Prof. Mor Peleg (University of Haifa)

Title: Collaboration between Government and Research Community to Respond to COVID-19: Israel’s Case

Mor Peleg 1,*, Amnon Reichman 2, Sivan Shachar 2, Tamir Gadot 1, Meytal Avgil Tsadok 3, Maya Azaria 4,
Orr Dunkelman 1,2, Shiri Hassid 3, Daniella Partem 4, Maya Shmailov 5, Elad Yom-Tov 6 and Roy Cohen 3

Triggered by the COVID-19 crisis, Israel’s Ministry of Health (MoH) held a virtual datathon based on deidentified governmental data. Organized by a multidisciplinary committee that included The University of Haifa’s Data Science Research Center and the Cyber Law and Policy Center, the Innovation Authority, and Microsoft Israel, Israel’s research community was invited to offer insights to help solve COVID-19 policy challenges. The Datathon was designed to develop operationalizable data-driven models to address COVID-19 health policy challenges. Specific relevant challenges were defined and diverse, reliable, up-to-date, deidentified governmental datasets were extracted and tested. Secure remote-access research environments were established. Registration was open to all citizens. Around a third of the applicants were accepted, and they were teamed to balance areas of expertise and represent all sectors of the community. Anonymous surveys for participants and mentors were distributed to assess usefulness and points for improvement and retention for future datathons. The Data-thon included 18 multidisciplinary teams (78 participants), mentored by 20 data scientists, 6 epidemiologists, 5 presentation mentors, and 12 judges. The insights developed by the three winning teams are currently considered by the MoH as potential data science methods relevant for national poli-cies. Based on participants’ feedback, the process for future data-driven regulatory responses for health crises was improved. Participants expressed increased trust in the MoH and readiness to work with the government on these or future projects. 


Speaker: Prof. Uri Hershberg (University of Haifa)

Title: The Artathon – Finding new Ways to accurately visualize complex biological data

The study of biology is at a turning point. Novel experimental methods make possible high-throughput imaging and molecular measurement at the single cell and single molecule level over thousands and millions of cells. Along with the great promise these technologies bring, they also call for new approaches to catalyze the assimilation of this overwhelming wealth of data. With new tools to measure biological dynamics and diversity new questions arise. Questions of process that take into account changes in individual molecules in the context of cell behavior and changes in individual cells in the context of systemic responses to physiological change. However, to even start to define these questions we need new ways to look at the data in its full richness. We need to visualize what we have measured. For the last 3 years in Philadelphia and Haifa (and now Porto) we have invited biology, computer science and design students to form interdisciplinary groups that together created new methods for the visualization of the changes in individually characterized Immune cell populations (http://artathon.stochasticity.org).
Our focus throughout has been to visualize B cell receptor (antibody) populations that can number in the millions for a single experiments. This is a special case of the more general question – How can we visualize millions of individuals? Which we divided into two sub questions: (i) How do we create an interactive visualization of the diversity of a population taking into account the different known characteristics of the individuals that build it? & (ii) How do we create an interactive visualization of the comparison of two populations?
For a more detailed description of our questions – http://clash.biomed.drexel.edu/artathon/).
This year, thanks to the pandemic, the Artathon was a hybrid events with small groups working together and in parallel on the east coast of the USA, in Porto in Portugal and here at the University of Haifa Namal Campus. The hybrid nature of the Artathon actually enhanced the collaboration and 4 interesting projects were built, two of which are still continuing one as a project in my lab and another as a proposed M.Sc. topic @INSEC TEC in Porto  (to see some of what wwas developed and conceptualized go here – https://github.com/Artathon2021 and to see the miro shared whiteboard we used to discuss things here – https://miro.com/app/board/o9J_lDLfORU=/.
In my lecture I will go over some of the insights we have gleaned from 3 Artathons (two supported by the DSRC) and our plans for future Artathons!

The Data Science Research Center of the University of Haifa supported 6 graduate students in 3-month internship programs at IBM and GE-Healthcare!

Two students worked at GE-Healthcare:
*Artur Shurin (Marine Technologies) developed a very accurate algorithm that can identify the exact location of caliper landmarks on an ultrasound images.
*Shunit Truzman (Marine Technologies) worked on Anomaly detection in structured clinical data (using HL7 FHIR standard) using Deep Learning approaches.

Four students worked at IBM:
*Yuli Zeira (Computer Science) worked on detecting escalation in the state of a chatbot-client dialogue at an early stage of the conversation, using classic machine learning methods ad state-of-the-art methods.
*Alex Kogan (Information Systems) used machine learning to identify possible future melanoma and kidney cancer patients.
*Roi Almakias (Statistics) worked on FreaAI — a product by IBM that automatically finds weaknesses in machine learning models. It is integrated into IBM services.
*Marcelo Feighelstein (Information Systems) built an AI-based application able to interpret flow charts on images and answer questions based on such interpretation.


248398999 877717432776451 7130526689456233812 n

Zoom Meeting Recording


Speaker: Ronit Marco (Reference Department, Younes & Soraya Nazarian Library, University of Haifa)

Title: The Academic Library – Tools and Services for Data Scientists

Abstract: Academic libraries tend to be perceived as stale establishments with shushing librarians and dust-collecting books on shelves. However, in the current Information Age, libraries are anything but that. Professional Information specialists continuously monitor the scientific information landscape to find new effective resources and tools that meet their patrons’ information needs. These are then presented to the researchers and students in online and offline personal or group guidance sessions.

The lecture will discuss the work of 21st century academic libraries and their interaction with their audiences, as well as resources and services that can benefit data scientists, including dataset repositories and innovative search engines.

 Dr. Tohar Dolev-Amit (Post doctorate student at the University of Haifa, on a Data Science Research Center Scholarship and a clinical intern at the AMCHA center in Haifa)

Title: Rupture Detection in Psychotherapy Using Human Action Recognition

AbstractAlliance ruptures are an integral part of psychotherapy and have the potential to either undermine treatment or enhance it. The most common method for identifying ruptures is through observational methods. However, it is labor intensive, with a long training and coding process. The direction of gaze and the facial expressions of the patient are a promising approach in psychotherapy research. We aim to identify whether ruptures are related to specific gaze direction and facial expression profiles.

The lecture will discuss the ability of using Human Action Recognition techniques, specifically Gaze Detection and Gaze Behavior Recognition as well as Facial expression through facial landmark detection, and facial action units’ recognition, in order to automatically detect alliance ruptures.

The third DSRC-students community meeting was held on Tuesday, 10.08.2021, at 16:00-17:00.

In this meeting we hosted Prof. Assaf Gottlieb, from the University of Texas Health Science Center at Houston (UTHealth), for a fascinating talk:

Title: Biomedical informatics approaches in the era of precision medicine

Abstract: Precision medicine, or personalized medicine, gradually making its way into medicine. The complex and high dimensional nature of clinical and genomic data presents a challenge to health providers but also an opportunity for biomedical informaticians.

Dr. Gottlieb is an Assistant Professor in the Center for Precision Health within the School of Biomedical Informatics, UTHealth, TX. In his talk, Dr. Gottlieb will present different approaches his lab employs to tackle various biomedical problems, integrating clinical and genomic data into a precision medicine machine learning framework and will discuss some data and implementation challenges.

Zoom Meeting Recording


Students Meeting 3 10.08.21

Dear all, 
We are happy to announce that this week we will host the AI4Biodiversity workshop in an hybrid mode – face-to-face and via zoom. 
For more details and registration please see the invitation file here
You are all invited!

Dear all, At Artathon 2021, an activity under the iReceptor + Project, we are devoted to find new ways to accurately visualize complex biological data.

If you are an artist who always dreamt to visualize the immune system

If you are a computer scientist who is interested in visualizations of high dimensional data

If you are a biologist who had enough of bar plots and pie charts

Then we need you, you would be delighted by the challenges we will post!

You should participate in this Artathon, 21-23 June (at the port of Haifa and via internet in Portugal and the US)

Please register here as soon as possible!



Full schedule here.


See you there,

Uri and DSRC Team

The International Workshop on AI Based Remote Patient Monitoring and Diagnosis in the Pandemic Age and Beyond (RPDM)

Tuesday 8:45-16:20 June 1, 2021

The RPDM was an online one-day workshop (see RPDM Webpage), that enabled cross-discipline computer scientists, who develop technologies and AI in tele-health to meet physicians who use telehealth as common practice as a measure to help individuals.   

Seven speakers participated in the RPDM workshop, and 98 individuals from Brazil, Italy, Switzerland and the UK, were registered to the workshop.  

The workshop was concluded with a panel on: Challenges in Developing International Unified Remote Therapieswhere the vision and obstacles were discussed.

This organized by Professor Hagit-Hel-Or and Doctor Shmuel Raz from the Computational Human behavior Lab. The workshop was hosted by the Dept of Computer Science at the University of Haifa and sponsored by the Data Science Research Center.

Participation was free but required registration. For further information including updated schedule please visit the conference web page:  RPDM Webpage, or contact us at razshmu@gmail.com

Links to the records of the lectures are here.



Zoom Meeting Recording

Ethics and Law in the Era of AI


Speaker: Prof. Sylvie Delacroix (University of Birmingham and The Alan Turing Institute)

Title: Bottom-up Data Trusts and Data Governance

Paper for your reference

Data Trusts Initiative


Speaker: Prof. Orna Rabinovich-Einy (University of Haifa)

Title: The Evolution of Online Dispute Resolution

Zoom Meeting Recording

Data Science Across Campus



14:00-14:20: Simon Korman (Computer Science Department) on Few-Shot Learning of Image Correspondence”

14:20-14:40: Tomer Sidi (DSRC) on Neural Machine Translation of Species to Species DNA”

14:40-15:00: Shani Stern (Sagol Department of Neurobiology) on “Using computational tools to predict drug responsiveness in psychiatric disorders”

15:00:15:20: Guy Avni (Computer Science Department) on Decision Making Using Graph Games”

15:20-15:40: Martin Mikl (Human Biology Department) on Using high-throughput assays to decipher the rules of gene regulation”

Zoom Meeting Recording  

The Synergy Between Data Science and environmental science

 Data Science of the Natural Environment: Opportunities and Challenges?
Prof. Gordon Blair (School of Computing & Communications, University of Lancaster)

Data science is the science of extracting meaning from potentially complex data. This is a fast moving field, drawing principles and techniques from a number of different disciplinary areas including computer science, statistics and complexity science. Data science is having a profound impact on a number of areas including commerce, health and smart cities. This will argue that data science can have an equal if not greater impact in the area of earth and environmental sciences, offering a rich tapestry of new techniques to support both a deeper understanding of the natural environment in all its complexities, as well as the development of well-founded mitigation and adaptation strategies in the face of climate change. In the talk, I will argue that data science for the natural environment brings about new challenges for data science, particularly around complexity, spatial and temporal reasoning, and managing uncertainty. I will conclude with a research roadmap highlighting ten top challenges of environmental data science and also an invitation to become part of an international community working collaboratively on these problems.

14:40-14:55: Questions


14:55-15:20: Harnessing eco-informatics to improve our understanding of ecological patterns and processes
Dr. Avi Bar-Massada (Department of Biology & Environment, University of Haifa)

Ecology strives to understand the interactions among species and their abiotic environment. A fundamental ecological question is what drives the geographic distribution of species, and to what extent these distributions are affected by biotic interactions and the abiotic environment. These questions are typically answered using species distribution models, which require large amounts of data on species occurrence locations as well as the environmental conditions in them. Such data requirements raise several challenges related to data acquisition and data analysis. Specifically, how can we improve data acquisition and data quality, and how can we analyze ever-increasing amounts of data without losing sight of the basic ecological principles that govern distribution patterns. These challenges are central to the emerging field of eco-informatics. In this presentation I will outline these issues, and exemplify, using three studies, the potential of data-science methods to enhance data collection, data analysis, and the ecological insights we derive from data. The first study deals with the automated classification of species occurrence in sites from data collected in the field. The second study presents a method to improve the predictive accuracy of species distribution models by the incorporation of data on multiple species at once. Finally, the third study reveals how environmental gradients drive species co-occurrence patterns across broad spatial extents. Together, these three examples highlight the potential of collecting and utilizing large ecological datasets in order to better understand ecological patterns and processes in a changing world.

15:20-15:27: Questions


15:27-15:53: GIScience integrated with computer vision for spatial examination of old visual sources
Dr. Motti Zohar (Department of Geography & Environmental Studies, University of Haifa)

Landscape reconstructions and deep maps are two major approaches in cultural heritage studies. In general, they require the use of historical visual sources such as maps, graphic artworks, and photographs presenting areal scenes and views, from which one can extract important spatial information. However, photographs, the most accurate and reliable source for scenery reconstruction, are available only from the second half of the 19th century onward. Thus, for earlier periods one can rely only on old artworks. Nevertheless, the accuracy and inclusiveness of old artworks are often questionable and must be verified carefully. In this talk, the use of GIScience methods with computer-vision capabilities to interrogate old engravings and drawings is presented. The developed approach is demonstrated using old depictions of Jerusalem that were created between the 17th and 19th centuries.

15:53-16:00 Questions


Zoom Meeting Recording

Speaker: Prof. Alexander Broadbent (Institute for the Future of Knowledge, University of Johannesburg)

Title: Can robots do epidemiology? Machine learning, causal inference, and predicting the outcomes of public health interventions

Abstract: This paper argues that machine learning (ML) and epidemiology are on collision course over causation. The discipline of epidemiology lays great emphasis on causation, while ML research does not. Some epidemiologists have proposed imposing what amounts to a causal constraint on ML in epidemiology, requiring it either to engage in causal inference or restrict itself to mere projection. We whittle down the issues to the question of whether causal knowledge is necessary for underwriting predictions about the outcomes of public health interventions. While there is great plausibility to the idea that it is, conviction that something is impossible does not by itself motivate a constraint to forbid trying. We disambiguate the possible motivations for such a constraint into definitional, metaphysical, epistemological, and pragmatic considerations, and argue that “Try it and see” is the outcome of each. We then argue that there are positive reasons to try, arising from the possibility of discovering meaningful new concepts of epidemiological and public health importance. Formal causal inference enforces existing classification schema prior to the testing of associational claims (causal or otherwise), but associations and classification schema are more plausibly discovered (rather than tested or justified) in a back-and-forth process of gaining reflective equilibrium. ML instantiates this kind of process, we argue, and thus offers the welcome prospect of uncovering meaningful new concepts in epidemiology and public health—provided it is not causally constrained.

A link to the paper


Speaker: Dr. Peter Dueben (Royal Society Research Fellow at the European Weather Center, ECMWF)

Title: Machine learning for numerical weather prediction

Abstract: The talk outlines how machine learning, and in particular deep learning, could help to improve weather predictions in the coming years and presents an overview of the work on machine learning methods that is ongoing at the European Centre for Medium-Range Weather Forecasts. Weather predictions are based on models of the Earth system — a huge system that consists of many individual components and shows chaotic behavior. For such a system, conventional tools are often struggling to provide satisfying results. On the other hand, a huge amount of data is available from both observations and modelling. Therefore, a large number of machine learning applications are currently tested in order to improve the different components across the workflow of numerical weather predictions. Whether these approaches will succeed is still unclear as there are also a number of challenges for the application of machine learning tools in weather predictions, such as the representation of multiscale atmospheric dynamics.

Zoom Meeting Recording

Human vs. Machines: Learning Traits of Artificial and Biological Neural Networks

 Prof. Edi Barkai (Sagol Department of Neurobiology, University of Haifa)

Title: A biophysical mechanism for acquisition and epigenetic inheritance of complex-learning skills

 Prof. Danny Keren (Department of Computer Science, University of Haifa)

Title: From linear regression to deep nets: a quick tour of machine learning


Zoom Meeting Recording

Computational social science

 Prof. Dirk Helbing (ETH Zurich)

Title: Simulating the World

Abstract: Given the Internet of Things, Big Data, and AI, can we now simulate the world? If yes, how should we go about it, and what could possibly go wrong? These are questions that can be of critical importance for the world, and actually for the future of all of us. 

 Dr. Michael Freedman (University of Haifa)

Title: Using Text as Data to Examine Religious Leader Rhetoric and Radicalization

Abstract: Why do some religious leaders adopt radical ideologies, while other do not? The ability to analyze very large amounts of texts with computational text analysis allows researchers to make progress on this important question. In this talk, I outline the two main statistical tools researchers use to analyze large amounts of text: (1) unsupervised topics models and (2) machine learning tools. I illustrate these tools using three related applications: the nationalist rhetoric of Israeli religious leaders during conflict with the Palestinians, the adoption of radical ideologies by militant groups in Afghanistan, and the turn to Jihad by some clerics

 Dr. Erez Shmueli (Tel-Aviv University, MIT Media Lab)

Title: Understanding, Predicting and Shaping Human Behaviour

Abstract: Big Data holds many promises, not only for the individual but also for the public good. At the individual level, Big Data can help users to become more connected, productive, and entertained. At the society level, Big Data creates tremendous opportunities in areas ranging from marketing to public health and urban planning. My work in this area has focused on utilizing data to build computational models of human behavior. Having such models would then allow us to better understand and predict future behaviors, and ultimately intervene when needed.  In this talk, I will focus on a concrete example from a recently published paper: “Money Drives: Can Monetary Incentives based on Real-Time Monitoring Improve Driving Behavior?”. This paper examines the effectiveness of monetary incentives based on real-time monitoring as means to improve driving behavior of professional drivers.


Zoom Meeting Recording Here

AstroInformatics + An Ignite-Talks Session

A New Approach to Periodicity Detection by Prof. Shay Zucker (Tel-Aviv University) 


A New Approach to Periodicity Detection by Prof. Shay Zucker (Tel-Aviv University)

In Astronomy, periodicity is pervasive in many forms: from rotation of asteroids,

through the orbits of binary stars, to stellar orbits around the Galactic Center. Thus,

detecting periodicity in astronomical data, which are often sparse and unevenly

sampled, has always been a staple of astronomical research. Astronomers use an

arsenal of methods to perform the task, none of them is perfect, and most of them

rely on some arbitrary assumptions or parameters. The most popular ones are

usually inspired by Fourier theory and essentially search for sinusoidal periodicities.

This talk will present a novel approach to periodicity detection – the Phase Distance

Correlation periodogram (PDC) – which is nonparametric, model independent and

computationally elegant. PDC is an application of some very recent developments in

statistics, and it opens up new horizons in the field of periodicity detection. It can

easily be extended to detect periodicities of new and unknown types, in various

modalities of data, not necessarily in Astronomy.

The talk will introduce the basic ideas of PDC, highlight its novelty, and demonstrate

its advantages in some types of data.

Cosmic sunsets and cosmic voids by Prof. Doron Chelouche
(Haifa Center 
for Theoretical Physics & Astrophysics)

On the use of constraints in statistical analysis by Prof. Ori Davidov
(Department of Statistics)

Geographical Roots of Cultural and Linguistic Traits by Dr. Assaf Sarid
(Department of Economics)

Ontology-based decision-support systems to enhance patients’ wellbeing by Prof. Mor Peleg
(Department of Information Systems)

A Seven Nation Consortium Using Healthcare Registry Data to Study 
Autism Spectrum Disorder by Prof. Stephen Levine (Department of Community Mental Health)

Data-driven research in information science by Prof. Daphne Raban
(School of Management, Faculty of Social Sciences)

Reverse Engineering the Prices of Spot Instances by Prof. Orr Dunkelman
(Department of Computer Science)


Zoom Meeting Recording Here

“Fake Folk Music” + An Ignite-Talks Session

14:00-14:40 – “Fake Folk Music” Prof. Oded Ben-Tal (Kingston University) 


I will be presenting on ongoing research collaboration using deep learning on melodic transcriptions of folk tunes. Our initial data set is a large collection of folk tunes crowd sources online. We used those to create a model that produces plausible outputs that share many of the characteristics of the original style. But as a composer, I wanted to prod this model in less conventional directions. Doing this, we discovered that the system did not in fact learn what we initially thought it did. My talk will explain our approach to machine learning with focus on evaluation of the model and the link between computational and creative research.

14:40-16:00 –
An Ignite-Talks Session – a series of ignite talks by leading experts from the University of Haifa on:

• Challenges and opportunities in time-to-event data by Dr. Bella Vakulenko-Lagun (Dept. of Statistics)

• From lone wolves to doves – inferring group behavior of pairs of museum visitors from sensors’ data by Prof. Tsvi Kuflik (Dept. of Information Systems)

• A reliability metric for complex data by Prof. Philip Tzvi Reiss

(Dept. of Statistics)

• Emergent properties of dynamic complex interactions by Dr. Osnat (Ossi) Mokryn (Dept. of Information Systems)

• Linking multiple domains of stress responsivity by Dr. Roee Admon

(School of Psychology)

• Statistical learning of dynamic systems by Dr. Itai Dattner (Dept. of Statistics)

• Text mining and information extraction: application to social media and across disciplines by Dr. Einat Minkov

(Dept. of Information Systems)

• AstroInformatics: cosmological sunsets by Prof. Doron Chelouche (Haifa Research Center for Theoretical Physics & Astrophysics)

Zoom Meeting Recording

The upcoming event will consist of a series of <10mins talks by DSRC members, aiming to foster collaborations, promote science, and also enjoy each other’s company. In addition, we hope that this gathering will promote multi/inter-disciplinary participation in the coming VATAT calls for funding in DS. To this end, our coming colloquium will be uniquely broad-scope and feature fascinating talks that cover a fair fraction of the science done at the DSRC. Specifically, the topics covered included :

· Information Security & Crypto

· Neural Networks

· Deep Learning and Applications

· Marine Sciences & Technology

· Biomedical Informatics

· Bioinformatics & Drug Design

· Online Communities

· Decision Making


• Eyal Privman (Evolutionary & Environmental Biology) – Genomic basis for the evolution of sociality in ants

• Uri Hertz (Cognition) Using cognitive models to study humans and machines

• Tomer Sagi (IS)- Why is integrating scientific datasets so hard and what can we do about it?

• Daniel Sher (Marine Biology) – Data Science, Oceans and Microbial Conversations

• Adi Akavia (CS) – Privacy-preserving machine learning training and prediction

• Roee Diamant (Marine Technologies) – Derivation and applications for the normalized matched filter

• Danny Keren (CS) – Silence is the essence of wisdom: reducing communication in distributed systems

• Judith Somekh (IS) – Methods in biomedical informatics to better understand modern diseases

• Rita Osadchy (CS) – Deep Learning & Applications

• Uri Hershberg (Human Biology) – The study of a immune cell populations as an example for the study of multi agent biological and social systems in the age of big data

• Ofer Arazy (IS) – Computational Social and Organization Science: Investigating Online Production Communities

• Mickey Kosloff (Human Biology) – Decoding the structural basis for specific protein-protein interactions – from cell signaling to corona.


We envision a vibrant meeting with ample opportunities to interact during the meeting via virtual “coffee breaks” using breakout rooms in zoom. Unfortunately, refreshments are on you this time….


05/03 between 11am and 2pm at the 600 pavilion of the main building, University of Haifa, Main Building.


יום הפוסטר הקהילתי למדעי הנתונים אורגן על ידי מרכז המחקר למדעי הנתונים, ב- 5 במרץ 2020.

הארוע התקיים בבניין הראשי ברחבת ה 600. 

חוקרים מכל המחלקות באוניברסיטה הוזמנו להציג מחקר שעוסק בניתוח נתונים מורכב.

בתחרות הוצגו מעל 35 פוסטרים ומחקרים חדשניים מובילים במדעי הנתונים מתחומים רבים.

מטרת היום הייתה לתת במה למחקר בין תחומי מתקדם העושה שימוש בשיטות מדעי נתונים כדי לנתח מערכי נתונים דיגיטליים, לעודד שיתופי פעולה בין חוקרים מתחומים שונים וליצור קהילת מדעי נתונים פעילה ואינטראקטיבית באוניברסיטת חיפה.

כדי לעודד סטודנטים וחוקרים להציג עבודה באיכות גבוהה, לדחוף למצוינות ולהבליט את המחקרים הטובים ביותר  הודענו כי התחרות נושאת פרסים כספיים ושלפוסטרים הטובים ביותר יוענקו פרסים כספיים. 

וועדה של 10 שופטים בכירים (בעיקר חוקרים ממחלקות שונות באוניברסיטה העוסקים במדעי הנתונים ובראשם פרופ’ מור פלג, פרופ’ גדי לנדאו, פרופ’ שולי וינטנר וחברי סגל נוספים, כמו גם מנכל”ית וסגן המחלקה הכלכלית של האוניברסיטה “כרמל” – גב’ אלקה ניר ועו”ד אייל ציוני, ונציגים מהתעשייה.

השופטים דרגו את הכרזות בהתבסס על חדשנות, חשיבות, עניין והשפעה. לאחר מכן נאספו התוצאות כדי לבחור את הפוסטרים המובילים.

בתום התחרות ובמעמד המשתתפים, הוכרזו הזוכים:

שני המקומות הראשונים ביום הפוסטרים של מרכז מחקר מדעי הנתונים של אוניברסיטת חיפה:
דוד גל מציוויליזציות ימיות: לפני אלפי שנים הפליגו אניות עם כיוון הרוח. הם יצאו מהאיזור שלנו אבל לא ברור איך חזרו. ניתוח נתונים רגיל לא מצא איך כי נראה שבאופן ממוצע באף חודש אין מספיר רוח. אבל מעבדים את נתונים נכון מבחינה סמנטית אפשר למצוא מספר הזדמנויות בחלק מחודי הקיץ. הנתונים הם נתונים מטרואולוגיים אמיתיים.
מקום ראשון נוסף לבבורה לווה, דריה אקיינק וטלי טרייביץ על אלגוריתם לעיבוד תמונה שמסיר את המים מתמונות שצולמו במעמקי הים


שני מקומות שניים ביום הפוסטרים של מדעי הנתונים באוניברסיטת חיפה:
אלברטו טסטולין ורועי דיאמנט פיתחו אלגוריתם שמוריד רעשים מנתונים אקוסטיים ומצליח לזהות גופים ולאכן אותם
אלכס קוגן מור פלג עירית הועברו וסמסון טו פיתחו אלגוריתם ויישמו מערכת שמתכננת טיפולים מבוססי תדריכים רפואיים נקיים מסתירות עבור חולים שיש להם מחלות מרובות (לכל חולה)


מקומות שלישיים ביום הפוסטרים של מרכז מחקר מדעי הנתונים של אוניברסיטת חיפה:

מגלי סגל ואילן שמשוני מפתחים אלגוריתם שמצליח להרכיב מחרסים שבורים של ממצאים ארכיאולוגים את השלם.
דרור קיפניס ורועי דיאמנט פיתחו אלגוריתם מורכב לזיהוי שריקות של ליוויתנים בודדים שמסתובבים בלהקה על סמך ההד שלהם וההרמוניה
חאנין קאראווני וסמירה אנדרסון ששאלו אם שיחזור מידע חושי ע”י מכשירי שמיעה אצל אנשים מבוגרים יכול גם לשפר את יכולת העיבוד החושית והקוגניטיבית של מערכת העצבים שלהם. ניתוח הנתונים שלהם הראה עדויות ששימוש בעזרי שמיעה מאט את ההידרדרות של מערכת העצבים שעוסקת בעיבוד מידע חושי ובקוגניציה.



For details and registration

27/01 14:00-16:00 University of Haifa, Education Building, room 570


Prof. Hagit Hel-Or (Department of Computer Science, U. of Haifa): Computer Vision for Human Behavior Understanding

Dr. Dan Levi (General Motors): Camera-based 3D Lane detection and other perception challenges in autonomous driving

Dr. Dan Feldman (Department of Computer Science, U. of Haifa):

Visual Navigation for Drones

Hagit Hel-Or
Title: Computer Vision for Human Behavior Understanding

Abstract :
In this talk I will present studies in which Computer Vision and Machine Learning are harnessed to study human behavior. We use various sensors and data capturing devices to study, body motion, hand motion, facial expression and emotion, and more. I will briefly review several studies performed in my lab, and will extend on one specific  study in which we developed an automated system to evaluate fall detection, using a novel multi-3d-camera system (work in collaboration with Prof Ilan Shimshoni and Physiotherapists at the Nehariya Hospital).

Dan Levi

Title: Camera-based 3D Lane detection and other perception challenges in autonomous driving

I will introduce “3D-LaneNet”, a network that directly predicts the 3D layout of lanes in a road scene from a single image. This work marks a first attempt to address this task with on-board sensing without assuming a known constant lane width or relying on pre-mapped environments. Our network architecture, 3D-LaneNet, applies two new concepts: intra-network inverse-perspective mapping (IPM) and anchor-based lane representation. The intra-network IPM projection facilitates a dual-representation information flow in both regular image-view and top-view. An anchor-per-column output representation enables our end-to-end approach which replaces common heuristics such as clustering and outlier rejection, casting lane estimation as an object detection problem. In addition, our approach explicitly handles complex situations such as lane merges and splits. Results are shown on two new 3D lane datasets, a synthetic and a real one. For comparison with existing methods, we test our approach on the image-only tuSimple lane detection benchmark, achieving performance competitive with state-of-the-art.

Dan Feldman 

Title: Visual Navigation for Drones

Abstract :
According to the law in Israel and US, you can use a drone inside the city or indoors only if its weight is less than 250 gram. Practically, this means that it can carry only a weak micro-computer and an RGB camera for autonomous navigation. This requires efficient real-time algorithms that do not exist today.

I will formalize some of these problems and suggest the first provably optimal and practical algorithms for some of them. This is by using modern optimization techniques such as core-sets, sketches, and Sum-Of-Squares (SOS). 
Demo videos on toy-drones inside and outside the lab will also be presented.
Joint work with Ibrahim Jubran, Alaa Malouf, and Yair Marom.
Based on paper in ICRA’19  and Outstanding Award paper in NeurIPS’19.

30/12 14:00-16:00 University of Haifa, Rabin Building, room 5015

Prof. Ilan Shimshoni (Department of Information Systems, University of Haifa)
Title: Solving Archeological Puzzles
Prof. Uzy Smilansky (Department of Physics of Complex System, Weizmann Institute of Science)
Title: Computer applications in Archaeology – Prospects, challenges and possible pitfalls
Dr. Moshe Lavee (Department of Jewish History, University of Haifa)
Title: From Digital Accessibility to Data Science: Ancient Hebrew Manuscripts as a Digital Humanities Test Case

Ilan Shimshoni

Ilan Shimshoni recieved his B.Sc. in mathematics from the Hebrew University in Jerusalem, his M.Sc. in computer science from the Weizmann Institute of Science, and his Ph.D. in computer science from the University of Illinois at Urbana Champaign (UIUC).
Ilan was a post-doctorate fellow at the faculty of computer science at the Technion, from 1995—1998, and was a member of the faculty of industrial engineering and management from 1998—2005. He joined the department of  Information Systems (IS) at Haifa University in October 2005. Ilan was the head of the IS department between 2005-2009.

Abstract :
This paper focuses on  the re-assembly of an archaeological artifact, given images of its fragments.
This problem can be considered as a special challenging case of puzzle solving.
The restricted case of re-assembly of a natural image from square pieces has been investigated extensively and was shown to be a difficult problem in its own right.
Likewise, the case of matching “clean” 2D polygons/splines based solely on their geometric properties has been studied.
But what if these ideal conditions do not hold?
This is the problem addressed in the paper.
Three unique characteristics of archaeological fragments make puzzle solving extremely difficult:

(1) The fragments are of general shape;

(2) They are abraded, especially at the boundaries (where the strongest cues for matching should exist);

and (3) The domain of valid transformations between the pieces is continuous.

The key contribution of this paper is a fully-automatic and general algorithm that addresses puzzle solving in this intriguing domain.
We show that our approach manages to correctly reassemble dozens of broken artifacts and frescoes.

Uzy Smilansky

Uzy Smilansky is a professor emeritus in the Department of Physics of Complex Systems at the Weizmann Institute of Science. He received his PhD at the Weizmann Institute of Science in 1969, and in 1971 he completed a post-doctoral fellowship at the Max Planck Institute in Germany. His research focused on the “fingerprints” of the classic chaos theory of quantum mechanics. He is also engaged in the development of computerized and mathematical methods for aiding archaeological research.
Smilansky’s academic career is highly interdisciplinary, covering a broad spectrum of fields and methods. He made pioneering contributions to various research directions: chaotic scattering, quantum billiards, scattering approach to quantization (the exterior interior duality). He introduced quantum chaos in graphs and made seminal contributions to the studies of nodal networks and domains.

In 2002 Smilansky started to develop computer based methods for archaeological research, concentrating on the taxonomy of small artefacts by their digital images in 3-D, obtained by precise optical scanning. This activity, now integrated in the Archaeology Institute at the Hebrew University, introduced methodologies which are used by most archaeology groups in Israel.

Abstract :
The application of computer based methods in archaeology increased exponentially during the past years. It  went hand in hand with similar growth in the availability of  relevant hardware and dedicated software, and with the intensification of the  interaction between computer scientists  and archaeologists. This process is far from being over, and I shall try to review  the prospects, challenges and the possible pitfalls, based on my experience gathered during the last decades

Moshe Lavee

Moshe Lavee is a senior lecturer in Talmud and Midrash in the department of Jewish History in the University of Haifa, director and founder of eLijah-Lab and co-director and founder of the new Haifa BSc program in Digital Humanities. He is also  the chair of the Inter-disciplinary Centre for Genizah Research in The University of Haifa.
In addition to his work in DH he studies Aggadic Midrash and Judeo-Arabic homilies in the communities of the Genizah, and the on themes of conversion, gender and the construction and demarcation of identity in rabbinic Literature, as well as literary forms, intermingling of genres and the role of authorship in Rabbinic and adjacent literatures. Moshe runs programs for young leadership and educators (“Mashavah Techila” and “Ruach Carmel”), working to foster relationships between the academic world and the larger community.

The Data Science Research Center announces a competitive poster day

Just DS it! Data Science Community in Action in University of Haifa

We invite researchers from all departments to present research that
involves complex data analysis.  The goal of the day is to highlight
interdisciplinary research that uses data science methods to analyze
digitized data sets, foster collaborations between researchers from
different fields, and create an active and interactive data science
community in University of Haifa.

The event will take place on Thursday March 5th, at the 600 pavilion
of the main building, between 11am and 2pm. 

The best posters, judged based on innovation, significance, interest
and impact will be given prizes (2000/1000/500 NIS for the first,
second and third respectively).

Refreshments will be served.

We hope to see all of you there. Your participation is important to
strengthen interdisciplinary research across the campus, and University of Haifa as an emerging force in data science.

Please register by February 15. Abstracts can be submitted by March 1.


יום הפוסטר הקהילתי למדעי הנתונים אורגן על ידי מרכז המחקר למדעי הנתונים, ב- 5 במרץ 2020.

הארוע התקיים בבניין הראשי ברחבת ה 600. 

חוקרים מכל המחלקות באוניברסיטה הוזמנו להציג מחקר שעוסק בניתוח נתונים מורכב.

בתחרות הוצגו מעל 35 פוסטרים ומחקרים חדשניים מובילים במדעי הנתונים מתחומים רבים.

מטרת היום הייתה לתת במה למחקר בין תחומי מתקדם העושה שימוש בשיטות מדעי נתונים כדי לנתח מערכי נתונים דיגיטליים, לעודד שיתופי פעולה בין חוקרים מתחומים שונים וליצור קהילת מדעי נתונים פעילה ואינטראקטיבית באוניברסיטת חיפה.

כדי לעודד סטודנטים וחוקרים להציג עבודה באיכות גבוהה, לדחוף למצוינות ולהבליט את המחקרים הטובים ביותר  הודענו כי התחרות נושאת פרסים כספיים ושלפוסטרים הטובים ביותר יוענקו פרסים כספיים. 

וועדה של 10 שופטים בכירים (בעיקר חוקרים ממחלקות שונות באוניברסיטה העוסקים במדעי הנתונים ובראשם פרופ’ מור פלג, פרופ’ גדי לנדאו, פרופ’ שולי וינטנר וחברי סגל נוספים, כמו גם מנכל”ית וסגן המחלקה הכלכלית של האוניברסיטה “כרמל” – גב’ אלקה ניר ועו”ד אייל ציוני, ונציגים מהתעשייה.

השופטים דרגו את הכרזות בהתבסס על חדשנות, חשיבות, עניין והשפעה. לאחר מכן נאספו התוצאות כדי לבחור את הפוסטרים המובילים.

בתום התחרות ובמעמד המשתתפים, הוכרזו הזוכים:

שני המקומות הראשונים ביום הפוסטרים של מרכז מחקר מדעי הנתונים של אוניברסיטת חיפה:
דוד גל מציוויליזציות ימיות: לפני אלפי שנים הפליגו אניות עם כיוון הרוח. הם יצאו מהאיזור שלנו אבל לא ברור איך חזרו. ניתוח נתונים רגיל לא מצא איך כי נראה שבאופן ממוצע באף חודש אין מספיר רוח. אבל מעבדים את נתונים נכון מבחינה סמנטית אפשר למצוא מספר הזדמנויות בחלק מחודי הקיץ. הנתונים הם נתונים מטרואולוגיים אמיתיים.
מקום ראשון נוסף לבבורה לווה, דריה אקיינק וטלי טרייביץ על אלגוריתם לעיבוד תמונה שמסיר את המים מתמונות שצולמו במעמקי הים.

שני מקומות שניים ביום הפוסטרים של מדעי הנתונים באוניברסיטת חיפה:
אלברטו טסטולין ורועי דיאמנט פיתחו אלגוריתם שמוריד רעשים מנתונים אקוסטיים ומצליח לזהות גופים ולאכן אותם
אלכס קוגן מור פלג עירית הועברו וסמסון טו פיתחו אלגוריתם ויישמו מערכת שמתכננת טיפולים מבוססי תדריכים רפואיים נקיים מסתירות עבור חולים שיש להם מחלות מרובות (לכל חולה).

מקומות שלישיים ביום הפוסטרים של מרכז מחקר מדעי הנתונים של אוניברסיטת חיפה:

מגלי סגל ואילן שמשוני מפתחים אלגוריתם שמצליח להרכיב מחרסים שבורים של ממצאים ארכיאולוגים את השלם.
דרור קיפניס ורועי דיאמנט פיתחו אלגוריתם מורכב לזיהוי שריקות של ליוויתנים בודדים שמסתובבים בלהקה על סמך ההד שלהם וההרמוניה
חאנין קאראווני וסמירה אנדרסון ששאלו אם שיחזור מידע חושי ע”י מכשירי שמיעה אצל אנשים מבוגרים יכול גם לשפר את יכולת העיבוד החושית והקוגניטיבית של מערכת העצבים שלהם. ניתוח הנתונים שלהם הראה עדויות ששימוש בעזרי שמיעה מאט את ההידרדרות של מערכת העצבים שעוסקת בעיבוד מידע חושי ובקוגניציה.

November 2019: Geoinformatics– 3 talks 

When: 25/11 14:00-16:00

Where: University of Haifa, Rabin Building, Room 5015


There is Plenty of Fish in the sea: Challenges in Underwater Computer Vision

Dr. Tali Treibitz (The Hatter Department of Marine Technologies, University of Haifa, Israel)

The ocean covers 70% of the earth surface, and influences almost every aspect in our life, such as climate, fuel, security, and food. The EU estimates that its maritime regions account for around 40% of its GDP. All over the world, depleting resources on land are encouraging increased human activity in the ocean, for example: gas drilling, desalination plants, port constructions, aquaculture, fish farming, producing bio-fuel, and more. These expanded activities influence the delicate ecology that is already threatened by global warming and ocean acidification, and present a risk of over-exploitation. The ocean is a complex, vast foreign environment that is hard to explore and therefore much about it is still unknown. Interestingly, only 5% of the ocean floor has been seen so far. As human access to most of the ocean is very limited, most of the operations in it rely on remote sensors. Thus, it introduces numerous research challenges in monitoring, surveying and data analysis on a wide scale.

In this talk I will cover some of our recent efforts in underwater computer vision and present open challenges.

Dr. Tali Treibitz is heading the Marine Imaging Lab, in the Department for Marine Technologies, Charney School of Marine Sciences, University of Haifa.
Previously: Dr. Tali Treibitz was a post-doc, working with David Kriegman in the Computer Vision group, Computer Science and Engineering department in the University of California, San Diego and with Jules Jaffe in the Jaffe laboratory for Underwater Imaging in the Scripps Institution of Oceanography.


Integrating multiple data sources for cross-disciplinary study of marine systems

Dr. Yoav Lehahn (The Strauss Department of Marine Geosciences, University of Haifa, Israel)

The study of the ocean is one of the biggest scientific challenges of the 21st century. It has a direct impact on our understanding of Earth’s climate and biogeochemical cycling, as well as on our ability to provide human society with food, chemicals and energy. Oceanographic research relies largely on in-situ and remotely-sensed observations, which describe physical, chemical and biological seawater properties at a given time and place. These observations are collected from various manned and unmanned platforms, including research vessels, floats, drifters, autonomous vehicles and satellites, providing abundance of interdisciplinary information on processes occurring over a wide range of spatial (from microns to thousands of km) and temporal (from seconds to decades) scales. Collection of oceanic data is tedious and costly. However, due to their wellrecognized importance, over the last century numerous in-situ and remotely-sensed measurements have been constantly performed in different parts of the World Ocean, resulting in the creation of a very large amount of oceanic data, scattered between different research domains and disciplines. In this talk I will discuss some of the possibilities, challenges and approaches associated with integration of multi-source data for improving our understanding on marine systems and their role in the Earth system.

In his work as an oceanographer Dr. Lehahn investigate the marine environment at the interface between scientific disciplines. For that he integrate theoretical work with acquisition, processing, analysis and interpretation of (i) in-situ data and (ii) remote sensing data from satellites and drones. His current work includes development of tools for high resolution remote sensing of the marine environment, implementation of artificial intelligence methods for oceanic data integration, and investigation of jellyfish and phytoplankton bloom dynamics off the Israeli coast of the Eastern Mediterranean

Dr, Yoav Lehahn studied Geophysics and biology in Tel Aviv University, and obtained a Ph.D. in Oceanography from Université Pierre et Marie Curie (University Paris VI) in Paris, France. He is a senior lecturer in the Department of Marine Geosciences in University of Haifa since 2017


Flood Forecasting in Data-Scarce Regions

Dr. Zvika Ben-Haim (Google LLC)

Floods are among the most common and most deadly natural disasters in the world. Flood forecasts are crucial for effective individual and governmental protective action. The vast majority of flood-related casualties occur in developing countries, where providing spatially accurate forecasts is a challenge due to scarcity of data and lack of funding. I will describe Google’s flood warning system, which provides flood extent forecast maps covering several flood-prone regions in India, and aims to eventually expand globally. To this end, we build high-resolution topographic maps from satellite data and run detailed hydraulic simulations to determine flood extent.

Dr. Zvika Ben-Haim received the B.Sc. degree in electrical engineering and the B.A. degree in physics in 2000, the M.Sc. in electrical engineering in 2005, and the Ph.D. degree in electrical engineering in 2010, all from the Technion—Israel Institute of Technology, Haifa, Israel. He is currently with the Google Israel R&D Center.


20191125 153353
WhatsApp Image 2019 12 02 at 07.52.07

14 November, 2019 @16 Hanamal St., Haifa: Talk by Intel’s AI group

The planned schedule is:

17:00-17:15       Introduction of the Data Science Roundtable/Meetup

17:15-18:00       Intel AI people (Industrial point of view)

18:00-18:30       Open Discussion “How can we promote collaboration between our research center and industry?” 

18:30-19:00       Planning of the next meeting and expectations from the Meetup series


28/10 14:00-16:00 University of Haifa, Rabin Building, room 5015

“Combatting Cancer through AI technologies: the case of breast cancer”

Michal Rosen-Zvi, PhD, Director, Healthcare informatics, IBM Research Labs, Visiting Professor, Faculty of Medicine, The Hebrew University

 In this talk I will review AI technologies that can be leveraged for combating cancer. I will focus on breast cancer and talk about gaps and challenges. I will share recent results of various teams aiming at detecting cancer in screening mammography images. I will end with a review of a deep learning algorithm that was trained on a linked dataset of mammograms and electronic health records and achieved breast cancer identification accuracy comparable to radiologists as defined by the Breast Cancer Surveillance Consortium benchmark for screening digital mammography. The algorithm has also revealed additional clinical risk features. The algorithm exhibits the potential to reduce the likelihood of breast cancer misdiagnosis. These results appear in a recent publication, see https://www.ncbi.nlm.nih.gov/pubmed/31210611.                        

Dr. Rosen-Zvi is the Director for Health Informatics at IBM Research and a visiting Professor at the Faculty of Medicine, the Hebrew University. She did PhD in computational physics at Bar-Ilan University and performed postdoctoral studies in Machine Learning at UC Berkeley and the Hebrew University. She is a co-author of more than 40 papers in peer reviewed venues and has five patents. She is also heading the Health Informatics Department at IBM Research, Haifa. She is a member of the Israeli National Council of Digital Health and Innovation at health services and she is co-leading a subcommittee on innovative and disruptive technologies.

 “Machine learning for drug analysis from real-world”

 Yishai Shimoni, PhD, Machine Learning for Healthcare and Life Sciences Group, IBM Haifa Research Labs

The increasing availability of healthcare data provides an opportunity to evaluate the effect of various interventions on various outcomes, such as disease-related clinical outcomes, adverse events, and cost. However, analysis of real-world data is notoriously biased, and requires carefully constructed machine learning methods to detect and correct these biases. In this talk I will briefly explain some of the biases that appear in the data, how we overcome them, and applications for such technology.

Dr. Shimoni is the manager of the Machine Learning for Healthcare and Life-Sciences group at IBM Research, Haifa. He has a PhD in Physics from the Hebrew University, and did his post doc at Columbia University in New York. He has over 20 publications and 3 patents, and has more than 15 years of experience in analyzing biological and clinical data and applying advanced computational methods to these data.

 “An integrative approach for pain assessment”

Pavel Goldstein, PhD,

School of Public Health, University of Haifa, Israel

Chronic pain affects 20% of the worldwide population, with costs greater than that of heart disease, cancer, and diabetes combined, and is the main cause of the opioid crisis. A critical challenge in pain management emerges from the fact that pain cannot be directly measured. Therefore, the current gold standard for the assessment of pain is self-reporting. However, Patient self-reports of pain are however influenced by subjective pain tolerance, memories of past painful episodes and current context, and their willingness and honesty in reporting pain level. Therefore, objective measures of pain are needed to better inform pain management.

In this talk, I will present my recent research that focuses on pain estimation in close to real-life settings through its expression across a range of channels — voice, and physiological/neural markers and body maps of multiple emotions. Particularly, I will present our mobile platform for tracking chronic pain patients that collects bodily representations of different emotions and pain, and audio/video records of pain narratives. In addition, I will present a wearable system, which objectively quantifies user’s pain perception based- on multiple physiological signals using sweep impedance profiling (SIP), electroencephalography (EEG), photoplethysmogram (PPG) and galvanic skin response (GSR). Based on the research, I am seeking to develop new tools for pain measurement and treatment.

Dr. Pavel Goldstein is a director of the Integrative Pain (iPain) Laboratory, located at the School of Public Health (University of Haifa) and a director of the Master Biostatistics track (in development). He has a Master’s degree in Biostatistics, a PhD in Social Neuroscience from the University of Haifa and did a postdoc at the University of Colorado Boulder. Dr. Goldstein has 15 publications and 1 patent and more than 10 years of data science consultations experience for industry and academia.



First Hackaton: Artathon

October 23-24, 2019

First Hackaton: Artathon

Organized by: Prof. Uri Herhberg, Department of Human Biology




Over twenty participants attended, from several Universities and from industry and public sector, including: The University of Porto, Portugal, Boston University, Shenkar College of Engineering, Design and Art, Bezalel Academy of Art and Design, Ort Braude, Faber Design and Branding, IBM research Haifa, and from the University of Haifa and Bar-Ilan University.


ביומיים האחרונים נערך Artathon מדעי הנתונים: פיתוח שיטות חדשות לויזואליזציה של נתונים ביולוגיים. היו סטים של נתונים של רצפי דנ”א של תאי B: תאי דם לבנים של המערכת החיסונית שמייצרים נוגדנים. תאים אלה מתחילים את ההתמיינות שלהם במח העצם אבל במהלך חיי האדם עוברים מוטציות רבות שמאפשרים למערכת החיסונית לייצר סוגים שונים של נוגדנים כנגד גופים זרים (או במקרה של מחלות אוטו-אימוניות, כלפי רקמות של האדם עצמו). Clone של תאים נוצר במקור מתא בודד שמתחלק במהירות רבה יותר מאשר תאים אחרים בגוף. אפשר ליצור עצים אבולוציוניים של כל צאצאי התא המקורי ולהראות את השונות בין התאים האלה. השונות היא במימדים רבים: הרצף הגנטי, היכולת שלו להיקשר לגופים זרים שונים, המיקום של התא – כלומר באיזו רקמה או איבר של האדם הוא נמצא: דם, מח, טחול, מעי וכדומה. חוקרים שונים רוצים לענות על שאלות שונות ולהבין את הצורה בה פועל המנגנון החיסוני והם רוצים להשוות בין עצים שמקורם בתא בודד שונה מאנשים שונים או מאותו אדם בנקודות זמן שונות, או מרקמות שונות. האם נוכל לחשוב על שיטות ויזואליזציה מעניינות שיעזרו לחוקרים להבין תוצאות של ניתוחים שמשווים בין העצים? האם נוכל לתכנת אב טיפוס שיקח נתונים כאלה ויציג אותם וזאת תוך 24 שעות?

זה היה נושא הארטאתון שאירגון פרופסור אורי הרשברג מהפקולטה לביולוגיה באוניברסיטת חיפה. זה גם ההאקתון הראשון של מרכז מחקר מדעי הנתונים של אוניברסיטת חיפה בניהולה של פרופסור מור פלג


 במקום הראשון בארטאתון קבוצת Clonebob עם אדיב אבו אליה , טום, מייקל פאבר ומייקל שהצמידו את התאים של ה-Clone ה clone לפי מיקומם באיברים בכל נקודת זמן המפה משתנה. גודל העיגולים מראה את מספר העותקים של תאים ששייכים לאותו צומת בעץ ה Clone



אתמול הסתיים הארטאתון של מדעי הנתונים בדגש ויזואליזציה של נתונים גדולים של התמיינות תאים של המערכת החיסונית – איך לתת תובנות לחוקרים ע”י ויזואליזציה. במקום השני אלכס קוגן ואורי הרשברג שמראים אילן יוחסין של תאים שמייצרים נוגדנים ומקורם בתא בודד. אלגוריתם בונה את העצים באלה אך יכולות להיות שגיאות בבניה או בנתונים ששימשו כמקור. הויזואליזציה נותנת מידע שמצביע על מהימנותו של כל צומת בעץ לפי 3 פרמטרים: מספר המוטציות מצומת האב (מספר שמופיע ליד כל צומת), מספר התאים באותה צומת (גודל הצומת) והקרבה לעלי העץ ובנוסף מאיזו נקודת זמן הצומת (צבע ירוק או כיול או אם בשתי נקודות הזמן אז בסגול)


המקום השלישי יאיר מרום, לירן איבנברג, ארין,  ואדמר אגויאר שפיתחו תוכנה שמשווה בין עצים לפי משתנים שהמשתמש בוחר. כאן מוצגים 3 משתנים שמאפיינים עצים: עומק העץ, מספר הבנים המקסימלי ומספר הבנים בעומק

22/10/2019 10:30 am to 11:30 am CRI, Education Building, room 570

“Liveness — a trampoline to ultimate technical agility”

Live programming is an idea espoused by programming environments from the earliest days of computing, such as Lisp machines, Logo, Hypercard, and Smalltalk. In common with all these systems is liveness — a feedback nearly instantaneous and evaluation always accessible.

In this talk, we will present undergoing research work aiming to bring liveness to more software development activities as a way to improve technical agility, so that software can be faster to write, visualize and understand, with the support of a few features of a Live Software Development Environment.

Bio: Ademar Aguiar is a Professor at Faculty of Engineering of University of Porto (FEUP) and researcher at INESC Porto, with over more than 20 years of experience on software development, software architecture and design (patterns, frameworks, infrastructures), agile methods, wikis, and open collaboration tools.

Current research interests include knowledge management practices and tools for software development teams and organizations (from code to documentations) using wiki-based philosophy and open collaboration tools.

Beyond the field of software engineering, Ademar is also exploring and applying Web 2.0 and social software to other audiences and fields, being presently the most important a social learning environment for schools and their communities (PLE/SLE).





היום התקיימה הרצאה ראשונה בחסות מרכז מחקר מדעי הנתונים של אוניברסיטת חיפה!

אדמאר אגויר מאוניברסיטת פורטו שבפורטוגל הרצה על איך אפשר לכתוב קוד כל כך אג׳ילי שהוא ממש חי! כלומר הקוד כולו פרמטרי והמשתמש יכול להביע את בחירתו מ סקאלה של אפשרויות והקוד מגיב מיידית לאינפוטים של המשתמש. ותוצר התוכנה מייד מראה את האפקט למשתמש, קצת כמו שאנחנו עובדים בגוגל דוק והשינוי מייד מתעדכן בענן. אבל זה לא מספיק, הקוד יוכל גם לצפות מה המשתמש כנראה ירצה לעשות ואילו דרישות חדשות תהיינה לו!

בהקשר של מדעי הנתונים, המשתמש יוכל להגיד מה בתוצאות חשוב שיהיה בולט וזה מייד ישתקף בהן! למשל, שיהיה ייצוג ויזואלי להבדלים הכי גדולים שיש בין קבוצות של עצמים, נניח בין תוכניות לימוד של מדעי הנתונים מול מדעי המחשב או בין תוכנית הלימודים שלנו מול זו של הטכניון (סתם דוגמאות)