Parsimonious User and Group Profiling in Venue Recommendation

This paper presents the University of Amsterdam’s participation in the TREC 2015 Contextual Suggestion Track. Creating e↵ective profiles for both users and contexts is the main key to build an e↵ective contextual suggestion system. To address these issues, we investigate building users’ and groups’ profiles for e↵ective contextual personalization and customization. Our main aim is to answer the questions: How to build a user-specific profile that penalizes terms having high probability in negative language models? Can parsimonious language models improve user and context profile’s e↵ectiveness? How to combine both models and benefit from both a contextual customization using contextual group profiles and a contextual personalization using users profiles? Our main findings are the following: First, although using parsimonious language model leads to a more compact language model as users’ profiles, the personalization performance is as good as using standard language models for building users’ profiles. Second, we extensively analyze e↵ectiveness of three di↵erent approaches in taking the negative profiles into account, which improves performance of contextual suggestion models that just uses positive profiles. Third, we learn an e↵ective model for contextual customization and analyze the importance of different contexts in contextual suggestion task. Finally, we propose a linear combination of contextual customization and personalization, which improves the performance of contextual suggestion using either contextual customization or personalization based on all the common used IR metrics.