Context information management using agents

Abstract

The thesis is about an effort taken towards achieving context awareness by efficiently managing context information. The management of context spans from defining context attributes to triggering of actions resulting from a context state change. The design of the model is targeted towards creating a context aware environment for an adhoc users/services collaboration system. Contextual data are grouped and stored as profiles and database tuples depending on context themselves. The context management model gathers profile and preferences of mobile users and services and monitors changes that the entities undergo. These changes are reflected as actions on the related entities depending on the context change. The thesis proposes a design, which enables creating relationships between context attributes and deriving semantics out of them to describe a situation. Once the situation is resolved, the model identifies the tasks to be carried out in such situation and triggers relevant actions. The scope of context management model is restricted to entities that are present in a physical room. (Abstract shortened by UMI.)