Tutorial 2

Title: Similarity Search,  Recommendation and Explainability over Graphs for different domains: Social Media, News, and Health Industry

Description

This tutorial offers a rich blend of theory and practice regarding graph mining algorithms, to deal with challenging issues such as scalability, data noise, and sparsity in recommender systems. We will provide a detailed theoretical mathematical background and a step-by-step analysis, by using an integrated toy example, which runs throughout all parts of the tutorial, helping the audience to understand clearly the differences among graph-based methods to infer similarity and provide recommendations in heterogeneous information networks. In particular, this tutorial surveys important research in a new family of recommender systems aimed at serving multi-dimensional social networks. We will provide the related work for similarity search on graphs. We will see the random walk-based algorithms (i.e., PageRank, SimRank, Katz, etc.) that can be used to provide contextual recommendations in multi-dimensional graphs, where there are many participating entities (users, locations, products, and the time dimension). Furthermore, we will present the time-evolving graphs which incorporate session nodes to model the time dimension. Moreover, we will present methods that use meta paths to infer similarity among entities and how these meta paths can be used for explaining either the similarity among entities or the suggested item recommendations. Moreover, we will present state-of-the-art graph neural networks, graph convolution networks, and graph embeddings (node2vec, metapath2vec, etc.) for similarity search and recommendation in graphs. Finally, we will demonstrate real-life systems that use the aforementioned graph-based algorithms for location-based social networks, news industry and the health domain along with user studies which were used to evaluate the acceptance of the users for these systems.

Authors’ Short Biography :

Panagiotis Symeonidis

University of the Aegean, Greece

Panagiotis Symeonidis is Associate Professor at the School of Information and Communication Systems Engineering, University of the Aegean, Greece. Before moving to Samos, he was Assistant Professor at the Faculty of Computer Science (scientific sector INF/01) of the Free University of Bolzano, Italy, till November 2020. Before moving to Bolzano he worked for 8 years as Adjunct Assistant Professor at the Department of Informatics of the Aristotle University of Thessaloniki, Greece. He received a B.Sc. degree in Applied Informatics from University of Macedonia at Thessaloniki in 1996. He also received a n M.Sc. degree in Information Systems from the same university in 2004. He received his Ph.D. in Web Mining and Information Retrieval for Personalization from the Department of Informatics of the Aristotle University of Thessaloniki in 2008. His research interests include web mining (usage mining, content mining and graph mining), information retrieval, collaborative filtering, recommender systems, social media in Web 2.0 and location-based social networks. He is co-author of 3 international books, 1 Greek book, 6 book chapters, 25 journal publications and 40 conference/workshop publications. His published papers have received more than 3000 citations according to google scholar. Recently, he recognized from AMiner among the Most Influential Researchers (https://www.aminer.cn/ai10/recommendation) in the last decade to the field of Recommender Systems. Half of his journal publications were published in top or highly ranked journals. One third of his conference publications have been published in top or highly ranked conferences. Lastly, he was enlisted in the top 2% of the most cited scientists in the world (https://www.unibz.it/en/news/136834-8-professor-innen-der-unibz-unter-top-2-der-wissenschaftler-innen-weltweit).

Duration : Half-Day (3 hours)