Abstract:  Tensors are multidimensional extensions of matrices that can represent multi-aspect/multi-modal data. Consider, for instance, a Knowledge Graph that captures relations between entities. This can be modeled as a three-mode (entity, entity, relation) tensor, where each frontal slice is a graph connecting two entities with respect to a certain relation. Tensor decomposition is a very powerful tool for analyzing and extracting hidden knowledge from such multi-aspect datasets.

There is a long list of applications where tensor methods have successfully enabled the integration of heterogeneous multi-aspect data and the embedding of all entities involved in the data into interpretable low-dimensional spaces, which can power a number of donwstream tasks from exploratory analysis to prediction and classification.

In this talk, we will explore the effectiveness of tensor decompositions in providing actionable insights, focusing on the following two applications:

(1) Graph Mining: We demonstrate how tensor decomposition can identify communities and their rich structure in graphs, produce node embeddings that perform on par with state-of-the-art, and instill robustness against adversarial attacks.

(2) Textual Insights: We focus on Misinformation on the Web, where we introduce the novel concept of tensor embeddings which achieve state of the art accuracy in misinformation detection with very limited amount of labels. Furthermore, we demonstrate the generality of our tensor embeddings in another very challenging NLP task of short text humor detection.

Bio: Evangelos (Vagelis) Papalexakis is an Associate Professor of the CSE Dept. at University of California Riverside. He received his PhD degree at the School of Computer Science at Carnegie Mellon University (CMU). Prior to CMU, he obtained his Diploma and MSc in Electronic & Computer Engineering at the Technical University of Crete, in Greece. Broadly, his research interests span the fields of Data Science, Machine Learning, Artificial Intelligence, and Signal Processing. 

His research involves designing interpretable models and scalable algorithms for extracting knowledge from large multi-aspect datasets, with specific emphasis on tensor factorization models, and applying those algorithms to a variety of real world problems, including detection of misinformation on the Web, explainable AI, and gravitational wave detection. 

His work has appeared in top-tier conferences and journals, and has attracted a number of distinctions, including the 2017 SIGKDD Dissertation Award (runner-up), a number of paper awards, the National Science Foundation CAREER award, and the 2021 IEEE DSAA Next Generation Data Scientist Award.