Social Analytics For Personalization In Work Environments

A user's context in work environments, or work context, provides fine-grained knowledge on the user's skills, projects, and collaborators. Such work context is valuable to personalize many web applications, such as search and various recommendation tasks. In this paper, we explore the use of work contexts derived from users' various online social activities, such as tagging and blogging, for personalization purposes. We describe a system for building user work context profiles, including methods for cleaning source data, integrating information from multiple sources, and performing semantic enrichment on user data. We have evaluated the quality of the created work-context profiles through simulations on personalizing two common web applications, namely tag recommendation and search, using real-world data collected from large-scale social systems.