Abstract:
Context Aware computing has been one of the most challenging and interesting developments from the past decade. The term context may be defined as, "Any information that can be used to characterize the situation of an entity. An entity is a person, place, or object that is considered relevant to the interaction between a user and an application, including the user and applications themselves."[3] Context awareness refers to the idea that computers can both sense, and react based on their environment. The importance of contextual information has been recognized by researchers and practitioners in many disciplines, including e-commerce personalization, information retrieval, ubiquitous and mobile computing, data mining, marketing, and management.
Along with context, the other topic of relevance to this work is that of recommender systems. Recommender Systems have been around for quite a while now with the most popular examples being that of Amazon[23] and Netflix that use collaborative filtering techniques to generate recommendations for their users. While a substantial amount of research has already been performed in the area of recommender systems, most existing approaches focus on recommending the most relevant items to users without taking into account any additional contextual information, such as time, location, or social circle etc. Despite some attempts being made at utilizing contextual information for generating recommendations, the problem remains largely unaddressed and tightly coupled with the base functionality of the service being provided.
In this work, we discuss how existing context aware systems exploit context and emphasize the relevance of this contextual information in recommender systems. We discuss the concepts of short-term and long-term context and how each of them can prove individually useful in the contextual pre-filtering and post-filtering processes of a context aware recommender system. We then, discuss the notion of social context and introduce a novel pre-filtering algorithm using collaborative filtering techniques which exploits a user's social context, and, provides a set of like-minded users to be used for generating recommendations. We will compare the performance of this algorithm with some existing techniques by evaluating the similarity of the set of users obtained in each case.
