Abstract: The emerging data science technologies in recent years has given rise to a new field of research consisting of context-aware query processing facilities in information systems. The extraction of timely actionable information from diverse data analysis is a real dilemma in data science. This study discusses a predictive analysis of personalization technique with quantitative user preference model. The first phase extracts personalized results from explicit learning. The second phase builds contextual preference rules form collection of personalized results using apriori algorithm. The view point of user interest retention and granular information processing examines the proposed personalization algorithm for user centric unification. Though many personalization algorithms have been proposed already they have limitations in terms of accuracy, user satisfaction and search time. The major advantages of the proposed system are reduced search time, improved customer satisfaction. Objective metrics, subjective user perception and behavioural measures are utilized to prove the quality of potentially effective result.