Enhanced Personalized Search using Social Data

Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 700–710, Austin, Texas, November 1-5, 2016. c⃝2016 Association for Computational Linguistics Enhanced Personalized Search using Social Data Dong Zhou1 , Séamus Lawless2 , Xuan Wu1 , Wenyu Zhao1, Jianxun Liu1 1. Key Laboratory of Knowledge Processing and Networked Manufacturing & School of Computer Science and Engineering, Hunan University of Science and Technology, Xiangtan, Hunan, 411201, China 2. ADAPT Centre, Knowledge and Date Engineering Group, School of Com- puter Science and Statistics, Trinity College Dublin, Dublin 2, Ireland dongzhou1979@hotmail.com, seamus.lawless@scss.tcd.ie Abstract Search personalization that considers the social dimension of the web has attracted a significant volume of research in recent years. A user pro- file is usually needed to represent a user’s inter- ests in order to tailor future searches. Previous research has typically constructed a profile sole- ly from a user’s usage information. When the user has only limited activities in the system, the effect of the user profile on search is also constrained. This research addresses the setting where a user has only a limited amount of usage information. We build enhanced user profiles from a set of annotations and resources that us- ers have marked, together with an external knowledge base constructed according to usage histories. We present two probabilistic latent topic models to simultaneously incorporate so- cial annotations, documents and the external knowledge base. Our web search strategy is achieved using personalized social query ex- pansion. We introduce a topical query expan- sion model to enhance the search by utilizing individual user profiles. The proposed ap- proaches have been intensively evaluated on a large public social annotation dataset. Results show that our models significantly outper- formed existing personalized query expansion methods which use user profiles solely built from past usage information in personalized search. 1 Introduction On today’s social web, users can enrich the social context of web pages. The most notable fact is that users can often freely tag web pages with an- notations (Gupta et al., 2011). These tags could be high quality descriptors of the web pages’ topics and a good indicator of web users’ interests. However, the uncontrolled manner of social tag- ging results in the use of an unrestricted vocabu- lary. This makes searching through the collection difficult and generally less accurate. Thus the so- cial annotation or bookmarking system demon- strates an extreme example of the vocabulary mismatch problem encountered in personalized web search. To tackle the problem, various per- sonalized query expansion (QE) and results re- ranking techniques have been proposed and eval- uated (Bouadjenek et al., 2016). There have been some attempts to achieve per- sonalized QE using social data. For example, Re- searchers have considered selecting the most re- lated tags from a user’s profile to expand queries (Bender et al., 2008; Bertier et al., 2009; Bouadjenek et al., 2011). Local analysis and co- occurrence based user profile representation have also been adopted to expand the query (Chirita et al., 2007; Biancalana et al., 2013). …

Content truncated to 500 words