Due to recent advancements in social networks, many people can consume diversified services on a daily basis and have developed an association with different communities of interest (COIs) via these services. However, a person only accesses a subset of these services at a given time either to consume certain services or to share information with a COI. This paper tries to answer two important research questions: 1) “how to dynamically capture user context from heterogeneous sources” and 2) “which services and COI are related to any given context.” To address these two challenges, we propose a framework called SenseFace, which provides user context from two sources: 1) a body sensor network (BSN) and 2) multimedia information contained the within social network space. We present the detailed design and implementation of the framework and share our preliminary test results.