Tutorial

Lifelong Learning Cross-domain Recommender Systems

Organizers: Dr Liang Hu, Dr Shoujin Wang, Dr Qi Zhang, and Dr Usman Naseem

Time & Location: 9:00 AM – 10:30 AM CDT at AT&T Hotel and Conference Center- Classroom #104

Contact Mail: Dr Liang Hu, Dr Shoujin Wang

Abstract

This tutorial presents the state-of-the-art and emerging studies on lifelong learning cross-domain recommender systems (RSs), including the latest and most advanced theories, methods, models, data, and applications. At the beginning of this tutorial, we will present the background and foundation of lifelong learning and RSs. Then, the 3C-principle, \emph{Complement, Composite, and Context}, will be utilized to build various lifelong cross-domain RSs by continuously fusing new knowledge across multiple domains. More specifically, we will respectively present the details of how to model lifelong learning cross-domain RSs with complementary, composite knowledge, and contextual knowledge over ever-evolving domains. 

Agenda

• Overview of lifelong learning cross-domain recommender systems 

• Lifelong cross-item domain recommender systems 

• Lifelong cross-user domain recommender systems 

• Lifelong cross-spatial domain recommender systems

• Lifelong cross-temporal domain recommender systems

• Lifelong cross-task domain recommender systems

• Summary

Slides

Download