AI in service of Humanity

Prof. Oliviero Stock is one of the top AI researchers worldwide in particular in the fields of intelligent user interfaces, applications in cultural heritage, and natural language processing. He received his Laurea in Mathematics from the University of Firenze and his “Specialization” (doctorate) in computer science from the University of Pisa in 1976. After serving as a researcher with the Italian National Council for Research in Rome, and coordinator of the National Strategic Project on Artificial Intelligence, he joined FBK-IRST Trento in 1987. At IRST, he founded the Natural Language Processing group, then the Communication and Cognitive Technology Division, and served as Director 1997-2001. Since then, he is a Senior Fellow and Head of Research in Artificial Intelligence.

Oliviero Stock initiated and directed a series of Italian–Israeli projects on AI and intelligent technologies between the University of Haifa and FBK-IRST (2003-2011), was one of the founders of the Trento node of the European Institute of Technology, and director of numerous European projects devoted to intelligent persuasion technologies, natural language, narrative negotiation, and ethical issues in AI. He is the author of two hundred and sixty published scientific papers, including over one hundred in refereed journals and books.

Prof. Stock is a Fellow of the European Association for Artificial Intelligence and its chair (1992-1994), Fellow of the American Association for Artificial Intelligence, and past President of the Association for Computational Linguistics (1996). He has been a member of the editorial board of a dozen scientific journals, and chair of two dozen international conferences or workshops.

As a tribute to his ongoing support and involvement in AI-related research, we are going to have a workshop focusing on AI in the Service of Humanity

The event will take place on May 27th, 10:30 - 16:40, at the Talia dormitories club (see the campus map below).

For any questions contact:

The secretariat of the Information Systems Department: +972 4 8288509 or

Prof. Tsvi Kuflik: +972 4 8288511; tsvikak@is.haifa.ac.il

Program

10:30 - gathering

11:00- 11:30 - Opening and greetings

Prof. Gustavo Mesch, rector of the university of Haifa

Prof. Eran Vigoda-Gadot, Dean of the Herta and Paul Amir faculty of Social Science

Prof Aaron Ben-Zeev - Former president of the university of Haifa

Prof. Martin Golumbic, Former head of CRI

11:30 - 12:45 - First session - Chairing Julia Sheidin: 3 speakers (25 min each)

Tsvi Kuflik -- “Enhancing the Museum Visit Experience - Ongoing Research Inspired by Prof. Oliviero Stock”

Claudia Goldman -- “Bridging the Human and Machine Communication Gap: Adaptive and Intuitive Interfaces”

Larry Manevitz -- “Small Data and Small Networks”

Meirav Hadad-Segev -- "Applying BDI-based mental model for Social Behavior in real-world applications"

13:10 - 14:30 - Lunch break

14:30 - 15:45 - second session - chairing Alan Wecker 4 speakers (25 min each)

Shuly Wintner -- “L1 Cognate Effects on L2 Lexical Choice: Computational Linguistics at the Service of Psycholinguistics”

Tamar Weiss -- “A Baker's Dozen: Lessons learned while collaborating with Prof. Oliviero Stock”

Paolo Traverso -- “Incremental Learning of Abstract Planning Domains from Continuous Perceptions”

15:50 - 16:00 - Dr. Stefano Ventura, the Scientific Attaché of the Embassy of Italy in Israel

16:00 - 16:15 - Prof. Ron Robin, president of the university of Haifa

16:15 - 16:40 - Oliviero Stock

Talks abstracts and speakers bio

Claudia Goldman - “Bridging the Human and Machine Communication Gap: Adaptive and Intuitive Interfaces”

Abstract: We frame the problem of understanding human machine interactions as one of bridging the gap in communication between an automated device and a human user. This gap results in non-natural or less-than- intuitive interactions and hence in a sub-optimal utilization of automation and a concomitant lack of understanding. The topic of HMI has primarily been studied by either focusing on the machine, or on the human side. Our claim is that we need to study the differences in communication between automated machines and human users to improve these interactions and maximize the benefits that can be attained by doing so. We present our computational approach to natural and intuitive interactions and conclude with applications to personal comfort in automotive case studies.

Bio: Claudia V. Goldman is a staff researcher in the User Experience Technologies Lab in General Motors Israeli Technical Center. She leads AI-UX projects in the automotive domain including smart mobility, intelligent planning and decision making, personalization and machine learning. Previously, she held researcher positions at Samsung R&D, University of Haifa, University of Massachusetts Amherst and Bar Ilan University. She holds a Ph.D. in Computer Science from the Hebrew University of Jerusalem. She has published over 70 publications in the Artificial Intelligence literature covering topics such as AI-based user machine interaction, multi-agent systems, coordination, decentralized control, and reasoning under uncertainty. She has 12 patents granted in AI, automotive and mobile interaction. She serves as a referee for the main Artificial Intelligence conferences and relevant journals

Meirav Hadad-Segev - "Applying BDI-based mental model for Social Behavior in real-world applications"

Abstract: In real-world applications, there is an increasing number of machines and robots which need to efficiently deliver tasks in a dynamic multi-agent environment. As such, they are required to deliver more advanced skills: (a) teamwork with other machines and/or with human beings; (b) adapting in real time to unplanned situations and/or changes; and (c) autonomous, collective decision-making. We present an innovative social AI platform. The platform models and implements a BDI-based mental model for Social Behavior. It enables machines and robots to manifest beliefs, desires, and intentions, and equally, attribute these mental states to others, with the understanding that the motivations of the other machines or humans might differ from their own. These social insights enable agents’ participation in mixed groups of both computerized and human agents through bestowing an awareness of other members of the group, awareness of the environment, cooperative decision making and autonomous interactions. The platform successfully applied to different industries such as gaming, simulation, and logistic robots.

Bio: Meirav Segev-Hadad is the founder and CEO of Brillianetor, a multi-agent AI platform that endows machines with the ability to interact socially within a group. She is an inventor of novel technologies in the area of Multi-Agent Systems and is also a multi-published author in the area of AI. Meirav has a proven record in senior management in the High-tech industry, specializing in the development of real-world applications for Artificial Intelligence in the fields of military systems, robotics, games, and simulators. Meirav holds a Ph.D. from Bar-Ilan University and Post Doctorate from University of Haifa with a specialization in Al Multi-Agent Systems. She also holds a variety of patents.

Tsvi Kuflik - “Enhancing the Museum Visit Experience - Ongoing Research Inspired by Prof. Oliviero Stock”

Abstract: The talks reflects on more that 15 years of research collaboration with Prof. Stock and his team, exploring the application of novel ICT technologies: From the joint German-Italian PEACH project through 8 years of a joint Italian - Israeli research project and until today's' research projects. Over the years, state of the art ICT technologies were demonstrated and experimented with, including natural language generation, automatic generation of multimedia presentations, indoor positioning, user modeling and personalization, interaction with smart devices and smart environments, augmented reality and more.

Bio: Prof. Tsvi Kuflik is the former chairperson of the Information Systems Dept. at The University of Haifa. and the founder and current co-chair of the Digital Humanities BSc. Program. Over the past ten years, the focus of his work was on ubiquitous user modeling applied to cultural heritage. In the course of his work, a “Living Lab” was developed at the University of Haifa including a museum visitors’ guide system for the Hecht museum. His main research interest is the use of intelligent user interfaces for ubiquitous computing within the “living lab”. In the lab, issues such as interaction with large, situated displays; interrupt management; navigation support; temporal and lifelong aspects of ubiquitous user modeling are studied. In recent years Tsvi has worked on applying his computer science knowledge to Digital Humanities, where he collaborates with Humanities researchers, e.g.. tasks that require text mining techniques for authorship attribution and automatic identification of rhetorical elements in poetry. Over the years, Tsvi has collaborated with numerous local and international researchers, supervised graduate students in this field, organized the PATCH workshops series (Personal Access To Cultural Heritage) and published over 200 scientific papers, 30 of them about this specific research area. A distinguished ACM scientist and a senior IEEE member, he received his BSc. and MSc. In computer science and his PhD. In information systems from Ben-Gurion University of the Negev, Israel.

Larry Manevitz - “Small Data and Small Networks”

Abstract: In recent years, applied neurocomputation has had remarkable successes using such techniques as "Deep Learning and Deep Networks" in conjunction with "Big Data". One of the hallmarks of Deep Learning is that the features can sometimes be extracted automatically as a byproduct of the classification training. However, in many very important situations, such as medical event related data, it is unrealistic to expect many data points. Nonetheless, in several recent works, together with my colleague Dr. Alex Frid, we have shown that subtle classifications can be extracted from e.g. Brain Data or Speech Data, using subtle feature selection techniques. For example, we can now classify the valence (i.e. "happy" or "sad") of personal memories from fMRI data obtained during free autobiographical recall; and similarly automatically classify the degree of Parkinson's disease solely and directly from a person's speech.

Bio: Larry Manevitz, Emeritus Professor and Head, Neurocomputation Laboratory

Paolo Traverso - “Incremental Learning of Abstract Planning Domains from Continuous Perceptions”

Abstract: We propose a framework for learning discrete deterministic planning domains. In this framework, an agent learns the domain by observing the action effects through continuous features that describe the state of the environment after the execution of each action. According to our approach, the agent learns its perception function, i.e., a probabilistic mapping between state variables and sensor data represented as a vector of continuous random variables called perception variables. We define an algorithm that updates the planning domain and the perception function. The algorithm can introduce new states, either by extending the possible values of state variables, or by weakening their constraints. It adapts the perception function to fit the observed data and adapts the transition function on the basis of the executed actions and the effects observed via the perception function. We define a measure of coherence of the learned model and prove the convergence to a coherent model.

Bio: Paolo Traverso is the Director of FBK ICT IRST at Fondazione Bruno Kessler - FBK

Tamar Weiss - “A Baker's Dozen: Lessons learned while collaborating with Prof. Oliviero Stock”

Abstract: This talk will review an extensive research program ensuing from more than a decade of collaboration with researchers from the FBK (Trento, Italy). The collaboration first began as part of the CRI-Haifa-Trento Research Agreement, and has led to multiple funded studies on diverse topics. Technologies were developed to enhance social interaction through story-telling and collaborative games based on Cognitive Behavioral Therapy for high-functioning children with autism. The results have demonstrated exciting new directions for harnessing the power of technology to improve key social impairments. We also developed a novel collocated technology to facilitate conflict escalation and de-escalation between Israeli-Jewish and Palestinian-Arab youth wherein face-to-face, tangible individual contributions were combined with joint actions. The results showed a significant shift to a more positive attitude toward a peer from another culture due to enhanced visibility of the conflict, and the notion that increased awareness enables a participant to deal with a dyadic cycle of conflict related actions and reactions. Most recently this work has been funded by Israeli and Italian Science and Technology Ministries to examine the ways in which collaborative technologies can reduce stereotypes between migrant and host citizens.

Bio: Prof. Weiss directs and manages clinical research projects at the University of Haifa’s Laboratory for Innovations in Rehabilitation Technology (LIRT) where she develops and evaluates novel virtual environments, haptic interfaces, co-located and online technologies to explore the effect of individual and collaborative rehabilitation. Rehabilitation and special education populations of interest include stroke, spinal cord injury, cerebral palsy, developmental coordination disorder, autism and head trauma. She worked with the Gertner Institute led by Prof. Mordechai Shani to develop and implement ReAbility Online, a tele-rehabilitation system which won first prize in the 2014 AbbVie-TEDMED competition for sustainable healthcare. Prof. Weiss’ research has been funded by the European Union, the Israel Science Foundation, the Israeli Ministry of Science and Technology and the Israeli Center of Research Excellence (I-CORE): Learning in a NetworKed Society. In this context, she studies collaborative technologies and 3D printing as novel pedagogical and rehabilitation methods. She is a founding board member of the International Society for Virtual Rehabilitation, steering committee chair of the International Conference on Virtual Rehabilitation series and was on the advisory board for the University of Southern California’s National Institute on Disability and Rehabilitation Research RERC on Technologies for Successful Aging with Disability. She has authored more than 200 peer-reviewed journal articles and book chapters, co-edited two books, and delivered numerous keynote addresses at international conferences.

Shuly Wintner - “L1 Cognate Effects on L2 Lexical Choice: Computational Linguistics at the Service of Psycholinguistics”

Abstract: We present a computational analysis of cognate effects on the spontaneous linguistic productions of advanced non-native speakers. Introducing a large corpus of highly competent non-native English speakers, and using a set of carefully selected lexical items, we show that the lexical choices of non-natives are affected by cognates in their native language. This effect is so powerful that we are able to reconstruct the phylogenetic language tree of the Indo-European language family solely from the frequencies of specific lexical items in the English of authors with various native languages. We quantitatively analyze non-native lexical choice, highlighting cognate facilitation as one of the important phenomena shaping the language of non-native speakers.

Bio: 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 Elect of the EACL.