Kapoor, Komal2020-09-022020-09-022013-11-18https://hdl.handle.net/11299/215937Computational models of preferences have been applied in various domains including economics, consumer research and marketing. They are also commonly used for designing recommender agents for suggesting new content to the users based on their inferred preferences. A major challenge for such systems is to cater to the changing needs of the users over time. Although, user preferences are known to be dynamic in nature, there are few methods for predicting these dynamics in a reliable way. In this thesis, the problem of defining predictive models of dynamic user preferences is addressed. A solution to this problem is provided by formulating a framework that incorporates history and time dependent changes in user preferences for items. Two types of changes in user preferences are identified. Firstly, user's interests are modeled as either favoring familiarity or looking for exploring new content. Secondly, user's preferences for familiar items are defined to change as a function of exposure for incorporating the psychological effects of boredom from repetition. Such a framework for estimating dynamic preferences of users provides unprecedented insights to user changing needs. These insights are proposed to be incorporated in solving two important problems for content services; user retention and temporally-aware recommendations.en-USReport