Title: Accurate and Trustworthy Recommender Systems: Algorithms and Findings

Date: Friday, April 12th, 2024

Time: 1:00 pm – 3:00 pm, EST

Location: CODA C1315 Grant Park / Zoom Link: https://gatech.zoom.us/j/9479542192?pwd=ck9HaXlsRFpKcWtpU1V5ODlkbHpBUT09

Name: Sejoon Oh

School of: Computational Science and Engineering

College of: College of Computing

Georgia Institute of Technology

Your website: https://sejoonoh.github.io/

Committee:

Dr. Srijan Kumar (Advisor, School of Computational Science and Engineering, College of Computing, Georgia Tech)

Dr. Chao Zhang (School of Computational Science and Engineering, College of Computing, Georgia Tech)

Dr. Constantine Dovrolis (School of Computer Science, College of Computing, Georgia Tech)

Dr. Julian McAuley (Computer Science Department, University of California San Diego)

Dr. Berk Ustun (Halıcıoglu Data Science Institute, University of California San Diego)

Abstract:

Modern recommender systems use deep learning algorithms trained with user-item interaction data to generate recommendations. However, current recommender systems still face diverse challenges with respect to accuracy, personalization, and robustness. In this thesis, we investigate such challenges and provide insights and solutions to them.

First, we study session-based recommender systems (SBRSs) and user intent-aware recommender systems, which have been proposed to enhance accuracy and personalization via modeling users’ short-term and evolving interests. We propose a novel SBRS: ISCON to assign precise and meaningful implicit contexts to sessions via node embedding and clustering algorithms. We also propose a new recommendation framework: IntentRec that predicts a user’s intent on Netflix and uses that as one of the input features of the next-item prediction of the user. Second, existing recommender systems cannot scale to large real-world recommendation datasets. To handle the scalability issue, we propose M2TREC, a metadata-aware multi-task Transformer model that uses only item attributes to learn item representations and is completely item-ID free. Finally, to address the extreme sparsity of recommendation datasets, we devise an influence-guided data augmentation technique DAIN that augments important data points for reducing training loss to the original data.

Apart from accuracy and personalization, we also analyze the robustness of existing recommender systems against input perturbations and devise a solution to enhance the robustness of the recommenders. We first introduce two Rank List Sensitivity (RLS) metrics that allow us to measure changes in recommendations against perturbations, and we propose two training data perturbation mechanisms (random and CASPER) for recommender systems. We further introduce a fine-tuning mechanism called FINEST that can stabilize predictions of sequential recommender systems against training data perturbations. Finally, we investigate the robustness of text-aware recommender systems against adversarial text rewriting. Our proposed text rewriting framework (ATR) can generate optimal product descriptions via two-phase fine-tuning of language models.