Τίτλος: Graph-based Methods for Recommender Systems
Ομιλητής: Αθανάσιος Νικολακόπουλος, Amazon
Ημερομηνία-χώρος: Παρασκευή 14 Μαΐου, 3-5μμ
Περίληψη: Item-based models are among the most popular collaborative filtering approaches for building recommender systems. Random walks can provide a powerful tool for harvesting the rich network of interactions captured within these models. They can exploit indirect relations between the items, mitigate the effects of sparsity, ensure wider itemspace coverage, as well as increase the diversity of recommendation lists. Their potential, however, can be hindered by the tendency of the walks to rapidly concentrate towards the central nodes of the graph, thereby significantly restricting the range of K-step distributions that can be exploited for personalized recommendations. In this talk, we will present RecWalk; a novel random walk-based framework that leverages the spectral properties of nearly uncoupled Markov chains to provably lift this limitation and prolong the influence of users’ past preferences on the successive steps of the walk—thereby allowing the walker to explore the underlying network more fruitfully. A comprehensive set of experiments on real-world datasets verify the theoretically predicted properties of the proposed approach and indicate that they are directly linked to significant improvements in top-n recommendation accuracy. They also highlight RecWalk’s potential in providing a framework for boosting the performance of item-based models. Building on the RecWalk framework we will also present PerDif; a novel method for learning personalized diffusions over item-to-item graphs for top-n recommendation. PerDif learns the teleportation probabilities of a time-inhomogeneous random walk with restarts capturing a user-specific underlying item exploration process. Such an approach can lead to significant improvements in recommendation accuracy, while also providing useful information about the users in the system. Per-user fitting can be performed in parallel and very efficiently even in large-scale settings. A comprehensive set of experiments on real-world datasets demonstrate the scalability as well as the qualitative merits of RecWalk and PerDif. Both methods achieve high recommendation accuracy, outperforming state-of-the-art competing approaches---including several recently proposed methods relying on deep neural networks.
Σχετικά με τον ομιλητή: Athanasios N. Nikolakopoulos is an Applied Scientist at Amazon, doing research on Machine Learning and Artificial Intelligence. Prior to joining Amazon, he spent 4 years as a Research Associate at the University of Minnesota, where he collaborated with the Digital Technology Center of the College of Science & Engineering (Recommender Systems and Graph-Based Learning), as well as the Department of Laboratory Medicine and Pathology at the Medical School (Computational Biology and Cancer Informatics). His research interests span the areas of Recommender Systems, Graph Mining, and Network Science, as well as Cancer Informatics. Within these fields, his work focuses on developing novel algorithms and models as well as useful software tools to tackle challenging real-world machine-learning/data-science problems. Dr. Nikolakopoulos has co-authored more than 30 papers, published in prestigious international conferences and journals. He is the co-recipient of 3 Best Paper Awards, for his work on Recommender Systems as well as Graph Learning Techniques. He serves on the program committee of several prestigious conferences on Data Science, Recommender Systems, and Machine Learning. Dr. Nikolakopoulos completed all his studies at the Computer Engineering & Informatics Department (CEID) at the University of Patras, where he earned his Engineering diploma in 2010, Master's degree in 2011, and PhD in 2016. He is a member of ACM, SIAM, and IEEE.