2^nd Workshop on Ubiquitous and Multi-domain User Modeling ( UMUM2023)

March 13 - 17 , 2023, HYBRID (ATLANTA, USA/ONLINE).

Scope & Topics

2^nd Workshop on Ubiquitous and Multi-domain User Modeling ( UMUM2023) shall be held from A March 13 - 17 , 2023, HYBRID (ATLANTA, USA/ONLINE). via hybrid mode. AI/ML technology has gradually shifted from a specific tool to a general tool for pervasive tasks. Typically, BERT, a well-known language pre-trained model has been successfully adapted to many language-related tasks. Recently, AI/ML has also been widely applied to user modeling for real-world applications such as context-aware recommendation. However, many studies focus on building user models only in one or few domains and applying them to identical domains. Thus, how to build ubiquitous and multi-domain user models and increase their applicability in real-world contexts remains an open problem. Here, we refer to “ubiquitous” as the capability to adapt to, and therefore exist in diverse contexts, and “multi-domain” as the capability of training jointly using data from multiple domains.

Topics of Interest :

• Deep Learning
• Transfer Learning
• Domain Adaptation; Active Learning
• Multi-task Learning
• Meta Learning
• Online Learning
• Continual Learning
• Representation Learning
• Knowledge Representation and Reasoning
• Context Modeling and Reasoning
• Spatio-temporal Modeling
• Activity Recognition
• Social Sensing and Modeling
• Cross-domain Recommendation
• Navigation
• Personalization

  Paper Submission

  Submit your contributions to Electronic Submission System; ( .pdf only).
  
  Note: If you are interested in special sessions, after submission via system, please mail this to mo-kurokawa@kddi-research.jp.

  Important Dates

  Submission Deadline	:	November 14, 2022
  Authors Notification	:	January 05, 2023
  Final Manuscript Due	:	January 30, 2023