Modelling the changing behaviour of a user is done by defining a user context for personalized services. Though many parameters have been identified to sense the context of a user automatically by a mobile device, identifying the right semantic context is still a challenge. Towards better defining user context for context aware mobile applications, this work proposes models for smarter context detection, a multidimensional evolutionary ontology, dynamic user profile generation, and a mapping of context and user preferences using a regression model. Some sample scenarios where the proposed method is an improvement over regular preference choices are shown. The proposed system is found to be dynamic and adaptive.
B. Rajathilagam, B, B., and Dr. Murali Rangarajan, “Dynamic Context-based User Profiles by Regression Modeling”, in of the 8th Annual International Conference on Computer Games, Multimedia and Allied Technology (CGAT 2015), GSTF,, Singapore, 2015.