Identification of Relevant Contextual Dimensions Using Regression Analysis

Abstract

The tremendous growth of information on the web has necessitated the need for recommendation systems. Although users' preferences used to vary under different situations which urge the requisite of context-aware recommendation systems. But the major issue to be addressed in context-aware recommendation systems is an efficient utilization of contextual dimensions, under which an item is consumed, are not equally important. Therefore, in this study, the determinants are analyzed that influences the user decision and their satisfaction towards watching movies. Thus a logistic regression model is developed to induce out the foremost factors that prevail the user satisfaction. The key findings of the study indicate that dominantEmo, endEmo, interaction, and weather are the most relevant contextual dimensions which integrated into the model would boost the performance of the model.