Title: Social Contextuality in Conversational Recommendation
Abstract: This research will explore new ways of supporting information seeking needs using advances in interactive recommendation and contextual informtion. Information seeking is becoming more and more difficult as users are increasingly aware of the ambiently information-rich environments they interact with while shopping, reading or browsing on the Internet. In order to accomodate this, recommender systems, generally seen to be good for finding items users probably want to see, need to be developed as more than suplimentary systems in online shops. This work will examine new forms of interaction similar to search-type relevance feedback and new approaches to the incorporation of detected context into recommenders. Given the intrinsically social nature of recommendation contextual cues and indicators will be used to determine social relationships. I will endeavour to discover means of improving the accuracy of recommendations offered as well as offering more efficient ways to browse collections of data using this context.