Coupling Everything: A Universal Guideline for Building State-of-The-Art Recommender Systems

Speakers: Liang Hu, Dr Shoujin Wang, Prof. Longbing Cao and Prof. Jian Cao

Authors: Liang Hu, Dr Shoujin Wang, Prof. Longbing Cao and Songlei Jian

Time: 08:30-12:00, SUNDAY AUGUST 11TH, 2019

Location: Naples 2701-2702, Venetian Macao Hotel Resort, Macao, China

Call for Paper: IEEE Intelligent Systems Special issue on Intelligent Recommendation with Advanced AI and Learning

Abstract:

Nowadays, the renaissance of artificial intelligence (AI) has attracted huge attention from every corner of the world. Specially, machine learning approaches have deeply involved in AI research in almost all areas, e.g., natural language processing (NLP), computer vision (CV) and planning. In particular, recommender systems (RS), as probably one of the most widely used AI systems, has integrated into every part of our daily life. In this AI age, state-of-the-art machine learning approaches, e.g. deep learning, have become the primary choice to model advanced RSs. It needs to note that various data are the power origin of RSs to conduct different recommendation tasks.

Classic RSs are built on the assumption that the relevant data, e.g. ratings and/or contents, are independent and identical distributed (IID), which suffers many issues, such as cold-start and data-sparse. Therefore, many state-of-the-art RSs are enhanced with machine learning techniques by incorporating different information. This tutorial will analyze data, challenges, and business needs in advanced recommendation problems, and take non-IID perspective to introduce recent advances in machine learning to model the next-generation RSs. This includes an overall of RS evolution and non-IIDness in recommendation, advanced machine learning for social RS, group-based RS, session-based RS, cross-domain RS, context-aware RS, multimodal RS, and multi-criteria RS in terms of model various couplings between users, items, contexts, modalities, and criteria.