Pr. Franco Alberto CARDILLO (CNR-Pisa, Italy)

Speech Title:"Large Language Models from Scratch: Theory, Architecture, and Practice? "

This lecture introduces the conceptual and mathematical foundations behind current Large Language Models (LLMs) and reconstructs their essential components step by step. We begin from first principles, such as linear and logistic regression, feed-forward networks, and back-propagation, to clarify how probabilistic learning and gradient-based optimization underpin neural text models. Building on this basis, we derive the Transformer architecture, showing how these principles come together in the model that forms the foundation of modern language generation systems.

Biography

Franco Alberto Cardillo is a permanent researcher at the Institute for Computational Linguistics (ILC) of the Italian National Research council. In 2007 he got a PhD degree in Computer Science from the University of Pisa defending a thesis titled “Attention-based Object Detection”. His research activity has always been focused on the study and the experimentation of Machine Learning and biologically-inspired models for the analysis of multimodal data. After joining ILC, he focused on the study of models and algorithms for text processing and analysis. Recently, he moved his attention spotlight on studying interpretable Artificial Intelligence models that merge symbolic (logic-based) and sub-symbolic (numerical) approaches that could partially overcome some major limitations of the two approaches taken on their own. His research activities include the study of innovative techniques for detecting rare anomalies within large datasets. He has been the scientific coordinator for ILC for several EU projects in Artificial Intelligence.

Pr. KHAN Fahad, CNR-Pisa, Italy

Speech Title:"An Introduction to Linguistic Linked Data"

In this course students will be introduced to the fundamentals of Linked Data and the Semantic Web, including the formal language RDF, the knowledge representation/ontology language OWL, and the query language SPARQL. They will also be introduced to the concept of the Linked Open Data cloud and will be shown how to query knowledge bases such as Wikidata and DBpedia. The course will focus on language resources such as lexicons as well as taxonomies and thesauri.

Pr. Madalina CROITORU (LIRMM – UM-Montpellier, France)

Speech Title:"Human - Intelligent Machines Interaction: An Interdisciplinary Perspective"

Pr. Nikola NIKOLOV, BDARG-ULIMERICK, Ireland

Speech Title : "The role of visualisation in the analysis of large and complex data"

Visualisation plays a central role in making sense of large and complex datasets. This talk explores two key strands: exploratory data analysis (EDA) through traditional and multi-dimensional plotting techniques, and network visualisation using force-directed and hierarchical layouts. Through selected examples, we will see how tools such as Plotly, D3.js, and Gephi support insight generation across different data types and domains. The talk highlights how visualisation can reveal structure, patterns, and anomalies that may remain hidden in raw data, and reflects briefly on its evolving role in modern data analysis workflows.

Biography

Dr Nikola S. Nikolov holds an integrated B.Sc./M.Sc. in Informatics from Sofia University (1995) and a Ph.D. in Computer Science from the University of Limerick (2002). He is an Associate Professor in the Department of Computer Science and Information Systems at UL, where he has been a faculty member since 2001. Dr Nikolov has published over 80 papers in areas including natural language processing, computer vision, network visualisation, and big data, with work appearing in Discrete Applied Mathematics, Journal of Big Data, and Journal of Medical Internet Research. He has served on academic and professional committees such as the IEEE Technical Committee on Visual Analytics and the programme committees for IEEE Big Data and IEEE CiSt. He has led research projects funded by Science Foundation Ireland and the Irish Research Council. Since 2006, Dr Nikolov has directed more than ten academic programmes at UL, including the M.Sc. in Artificial Intelligence and Machine Learning and the B.Sc. in Computer Systems. He has played a key role in shaping data science and AI education at UL, contributing to the development of the online National M.Sc. in AI, the M.Sc. in Business Analytics, and several skill-conversion programmes

Pr. Salem BEN FERHAT, CRIL, Université d'Artrois, CNRS, France

Speech Title:"Tracking Conflicts: From Detection to Resolution in Description Logic"

Pr. Carole DELENNE, AMU-Marseille,France

Speech Title:"River flood simulation using Graph Neural Network"

Professor Ahlam BEGDOURI, FST, USMBA, Fez Morocco

Speech Title:"From human reasoning to Case-Based Reasoning: Case studies"

Case-Based Reasoning (CBR) is a decision-making paradigm that relies on the reuse of past experiences to address new problems. Its foundational principle is that similar problems tend to have similar solutions. In practice, CBR systems store previously solved problems as cases in a case library, which can later be retrieved and adapted to solve new situations.

In this talk, the audience will be introduced to the origins and historical development of the CBR approach, as well as to the fundamental concepts underlying its reasoning cycle. Several case studies from diverse application domains will be presented to illustrate how data are modeled, similarity measures are defined, and how case retrieval and retention are performed.

Biography

Ahlame BEGDOURI is professor in computer sciences at the Faculy of Sciences and Technology of Fez (FST-Fez), University Sidi Mohamed Ben Abdellah (USMBA), since 1996. She held the position of Vice-Dean in charge of research and international cooperation at the FSTS (2021-2024) and is currently the head of the USMBA Doctoral Center of Sciences and Technologies and Medical Sciences since 2024. Her research topics include context-aware and adaptive applications, Case-Based Resoning, Communities of Practice, social and mobile learning. She is cothor of more than 30 research papers. She was involved in several funded research projects as coordinator of the FP7 MoICT project (2011-2014), local coordinator of AUF-IFFADEM DISCOMOB project (2013-2015) and local coordinator of the Horizon Europe STARWARS project (2023-2026).

Professor Ali IDRI (ENSIAS-Rabat, Morocco)

Speech Title:"Ensemble learning: Concepts, Techniques, and Applications"

Professor Azzedine ZAHI, LSIA, FST, USMBA, Fez Morocco

Speech Title:"Bridging semantics and learning : embedding from text to graphs"