The talk “Evaluating Diversification in Group Recommendation of Points of Interest” will be presented November 21^st, at 15:00, room I-105, moderated by Prof. Rosaldo Rossetti (DEI).

 Abstract:

With the massive availability and use of the Internet, the search for Points of Interest (POI) is becoming an arduous task. POI Recommendation Systems have, therefore, emerged to help users search for and discover relevant POIs based on their preferences and behaviors. These systems combine different information sources and present numerous research challenges and questions. POI recommender systems traditionally focused on providing recommendations to individual users based on their preferences and behaviors. However, there is an increasing need to recommend POIs to groups of users rather than just individuals. People often visit POIs together in groups rather than alone. Thus, some studies indicate that the further users travel, the less relevant the POIs are to them. In addition, the recommendations belong to the same category, without diversity. This work proposes a POI Recommendation System for a group using a diversity algorithm based on members’ preferences and their locations. The evaluation of the proposal involved both online and offline experiments. Accuracy metrics were used in the evaluation, and it was observed that the level at which the results were analyzed was relevant. For the top 3, recommendations without diversity performed better, but diversification positively impacted the results at the top 5 and 10 levels.

 About the Speaker:

Frederico Araújo Durão is an Associate Professor at the Institute of Computing of the Federal University of Bahia. Frederico Durão did his post-doctoral research at Insight Centre for Data Analysis, University College Cork, Ireland in 2016/2017. In 2012, he obtained his PhD in Computer Science from the University of Aalborg, Denmark. Frederico Durão has reviewed and published several articles in conferences and journals relevant to the areas of Information Systems, Recommender Systems, and Semantic Web. Currently is a senior researcher and the project leader of the RecSys Research Group in Brazil.

The talk ““Insert Coin” – A long-term study of education gamification” will be presented December 9^th, at 3pm, room I-105, moderated by Prof. Daniel Mendes (DEI).

Abstract:

Education nowadays still follows, for the most part, the traditional lecture-based teaching paradigm that has been the leading approach for well over a century. This flies in the face of current personal learning dynamics, in a world where information is increasingly at our fingertips. This mismatch between student expectations and classroom practice directly impacts their interest, engagement, and will to learn. Gamification has shown promise, in recent years, as a way to bring a game-like experience to several contexts, including education. Using it, learning becomes a game, with expected increases in motivation and, consequently, learning outcomes. Over a period of thirteen years we have gamified a MSc-level course, Multimedia Content Production. We tried to appeal to student’s nature as gamers and provide a flexible experience whereby they can exercise their autonomy. We will present how the game experience has evolved over that period of time and the lessons learned based on student expectations and reactions in this context. What is more, it soon became clear to us that students do not all react to the gamified experience in the same way. We can profile them using a four-cluster taxonomy, that has shown resilience throughout the years and that serves as the basis for an adaptive learning experience that will, finally, allow us to depart from the monolithic one-size-fits-all approach to education.

About the Speaker:

Daniel Gonçalves is full professor at the Computer Science Department of Instituto Superior Técnico – University of Lisbon, and a researcher in the Graphics and Interaction area at INESC-ID, where he specializes in Human-Computer Interaction (HCI), in particular in Education Gamification and Information Visualization. With a prolific academic output, he has authored over 200 scientific articles and a textbook on HCI, guided 11 doctoral and 100+ master’s students, and played a prominent role in various research projects in the area.

The lecture “Design and AI Innovation” will be presented on October 24, at 11:00, in INESC TEC’s Auditorium B.

Abstract:

“As early as 2011, Marc Andreesen identified that the world was facing a broad technological and economic shift in which software companies were poised to command much of the world’s economy. Now, 13 years later, the emergence of computing, data, and AI have impacted all industries. In this talk, I will examine how AI is changing my discipline, design, but also how design is changing AI. I will reflect on these ideas along with the emergence and rapid growth of generative AI and Large Language Models. I will identify new spaces for product innovation that utilize the most fruitful elements of the practice of design and AI as a design material.”

 Bio:

Jodi Forlizzi is the Herbert A. Simon Professor of Computer Science and Human-Computer Interaction in the School of Computer Science at Carnegie Mellon University. She is also the Associate Dean of Diversity, Equity, and Inclusion in the School of Computer Science. Jodi has advocated for design research in all forms, mentoring peers, colleagues, and students in its structure and execution, and today it is an important part of the HCI community. Jodi studies the ethical impacts of human interaction with AI systems in front-line service industries including healthcare and hospitality. She also develops methods and tools to ensure that product developers can mitigate ethical harms and bias during product development. Jodi is an ACM SIGCHI Fellow and recently received its Lifetime Research Award. She recently testified to the US Senate in one an AI Innovation Briefing and is a central advisor to the AFL-CIO Tech Institute regarding technology research.

“Historically, the run time of implicit mechanics has been dominated by the time required to solve a large sparse linear system. The default solver is a multifrontal sparse matrix factorization, which will reliably solve ill conditioned, indefinite problems. The multifrontal method turns a sparse matrix factorization into a directed acyclic graph of smaller, dense “frontal” matrix factorizations, and these can be accelerated using Graphics Processing Units. As the number of processors used grows into the thousands, reordering the sparse matrix to reduce the storage and operations required to factor it, is the emerging computational bottleneck. Reordering is NP-complete, and in computational mechanics the preferred heuristic is nested dissection, i.e., recursive graph partitioning. Finding a balanced min cut is NP-hard, and classical codes such as ParMetis have limited parallel scaling. This talk will also discuss on-going work to explore a new generation of specialized devices for solving optimization problems. These include the D-Wave adiabatic quantum annealer, so called Silicon annealers produced by Fujitsu and Toshiba, the LightSolver Laser Processing Unit. The Digital Annealer is a dedicated chip that uses non-von Neumann architecture to minimize data movement in solving combinatorial optimization problems.”

“Accelerating Implicit Mechanics” will be presented October 10, at 15:00, room Vasco Sá (L119) – Sala de Atos do Departamento de Engenharia Mecânica.

“Dr. Robert F. Lucas received his BSc, MSc, and PhD degrees in Electrical Engineering from Stanford University in 1980, 1983, and 1988 respectively. He is currently an Ansys Fellow where he is responsible for the default multifrontal linear solver used in LS-DYNA and MAPDL. Previously, he was the Operational Director of the University of Southern California (USC) – Lockheed Martin Quantum Computing Center. Before joining USC, he was the Head of the High-Performance Computing Research Department in the National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory and before that the Deputy Director of DARPA’s Information Technology Office. From 1988 to 1998 he was a member of the research staff of the Institute for Defense Analyses’s Center for Computing Sciences. From 1979 to 1984 he was a member of the Technical Staff of the Hughes Aircraft Company.”

Note: This talk is preceded by another talk, geared towards Mechanical Engineering and focusing on the use of ANSYS/LS-DYNA for modeling and simulation, by the same speaker at 14:00, in the same room, entitled “An Industrial Grand Challenge”. You are all welcome.

“Machine learning (ML), particularly deep learning, is being used everywhere. However, not always is used well, ethically and scientifically. In this talk we first do a deep dive in the limitations of supervised ML and data, its key component. We cover small data, datification, bias, predictive optimization issues, evaluating success instead of harm, and pseudoscience, among other problems.  The last part is about our own limitations using ML, including different types of human incompetence: cognitive biases, unethical applications, no administrative competence, copyright violations, misinformation, and the impact on mental health. In the final part we discuss regulation on the use of AI and responsible AI principles, that can mitigate the problems outlined above.”

“The Limitations of Data, Machine Learning & Us” will be presented September 10, at 11:00, room B032. Free entry but registration required here.

“Ricardo Baeza-Yates is the Director of Research at the Institute for Experiential AI of Northeastern University, as well as part-time professor at the Dept. of Computer Science of University of Chile. Before, he was VP of Research at Yahoo Labs, based in Barcelona, Spain, and later in Sunnyvale, California, from 2006 to 2016. He is co-author of the best-seller Modern Information Retrieval textbook published by Addison-Wesley in 1999 and 2011 (2nd ed), that won the ASIST 2012 Book of the Year award. From 2002 to 2004 he was elected to the Board of Governors of the IEEE Computer Society and between 2012 and 2016 was elected for the ACM Council. In 2009 he was named ACM Fellow and in 2011 IEEE Fellow, among other awards and distinctions. He obtained a Ph.D. in CS from the University of Waterloo, Canada, and his areas of expertise are responsible AI, web search and data mining plus data science and algorithms in general.”

“The acquisition of features for patient diagnostics, treatment planing and monitoring purposes is costly. Moreover, when patients with chronic diseases are called to used mobile health apps, they are also called to interact with the app in a regular way; the willingness to do so may wane with time. In this talk, we see forms of missingness in data collected in a clinic for treatment planning and in data collected with an app for monitoring. Then, we discuss methods that iteratively build up a minimal feature subspace for treatment outcome prediction, and neighbourhood-based methods that build up a minimal data space for patient condition monitoring. The methods have been applied on clinical data of tinnitus patients and on mhealth data of patients with tinnitus or diabetes. The results demonstrate that small subsets of features are often adequate for prediction.”

“Temporal mining on systematically sparse medical data” will be presented July 22, 15:30, room B012. The talk will be moderated by João Moreira (DEI).

Myra Spiliopoulou is Professor of Business Information Systems at the Faculty of Computer Science, Otto-von-Guericke-University Magdeburg, Germany. Her main research is on mining temporal complex data and extracting predictive patterns from evolving objects. One of the core application areas for her research, and a constant source of inspiration is health: her work encompasses methods and findings from observational medical data, from clinical studies, from digital health solutions, and from experiments on understanding the process of human and animal learning. She is involved as (senior) reviewer in major conferences on data mining and knowledge discovery, as Action Editor in the Data Mining and Knowledge Discovery Journal of Springer Nature, as Special Editor for survey papers in the International Journal of Data Science and Analytics (JDSA) and as Editorial Board Member for the Artificial Intelligence in Medicine Journal. In 2016, 2019 and 2023, she served as a PC Chair of the IEEE Int. Symposium on Computer-Based Medical Systems (CBMS). In 2024, she serves as senior reviewer for KDD 2024. She also serves as one of the Journal Track Chairs for ECML PKDD 2024, responsible for the submissions to the Machine Learning Journal. In May 2023, she received the Distinguished Service Contributions Award for the Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD).

“The talk will discuss Concept Relevance Propagation (CRP) and Prototypical Concept Explanation (PCX), two next-generation Explainable AI (XAI) methods, which explain individual predictions of an AI model in terms of human understandable concepts and allow to systematically investigate global model behaviors. Furthermore, the talk will present the potential of these novel methods to provide deep insights into the representation and reasoning processes of LLMs.”

 “Towards Next-Generation Explainable AI” will be presented July 16, 15:50, in room B032, moderated by Prof. Henrique Lopes Cardoso (DEI).

“Wojciech Samek is a professor in the EECS Department at the Technical University of Berlin and is jointly heading the Department of Artificial Intelligence at Fraunhofer Heinrich Hertz Institute (HHI), Berlin, Germany. He is Fellow at the BIFOLD – Berlin Institute for the Foundation of Learning and Data as well as the ELLIS Unit Berlin. He has co-authored more than 200 peer-reviewed journal and conference papers, and has received multiple best paper awards for his work in the field of XAI.”

What is the point of play and games when the climate catastrophe is looming? In this talk the speaker will explore the interconnections between the culture that led us to the climate catastrophe, and will reflect about the future of play and games after the end of a world.

“Games and Play and the End of a World” will be presented June 20, at 10:30, in room B011, moderated by Prof. António Coelho (DEI).

“Miguel Sicart is a Professor at the Center for Digital Play (digitalplay.itu.dk), IT University of Copenhagen. He has been researching games and play since the early 2000s, focusing on the intersection of games, ethics, and play. He is the author of, among others, Play Matters and Playing Software (The MIT Press, 2014, 2023). He is currently working on ridiculous software and the poetics of roguelikes.”

“Natural language inference (recognizing textual entailment task in the past) belongs to the most prominent tasks in current NLP, it is a keystone of natural language understanding. NLI can be stated as a classification task whether a given hypothesis can be inferred from a  given premise. In this talk, we are going introduce a large variety of tasks accompanied by illustrative examples and review corresponding state-of-the-art results. This talk may serve as starting point for anyone who want to apply new approaches and models to investigate these not so much known tasks (as well as apply them in downstream applications).”

“A Survey of Tasks Derived from or Related to Natural Language Inference” will be presented February 23, at 17:00, room I 025, moderated by Prof. Carlos Soares (DEI).

“Martin Víta graduated at the Faculty of Mathematics and Physics, Charles University in Prague in the field of discrete models and algorithms. Later, he obtained PhD degree at the Faculty of Informatics, Masaryk University in Brno in natural language processing. Currently, he serves as an assistant professor at the Faculty of Informatics and Statistics, Prague University of Economics and Business where he teaches mathematics and text analytics. He is also a researcher at Czech Academy of Sciences where he focuses in ML and text mining topics.”