A Language Modeling Approach to Personalized Search based on Users’ Microblog Behavior Arjumand Younus 12 , Colm O’Riordan 1 , and Gabriella Pasi 2 1 Computational Intelligence Research Group, Information Technology, National University of Ireland, Galway, Ireland 2 Information Retrieval Lab, Informatics, Systems and Communication, University of Milan Bicocca, Milan, Italy {arjumand.younus@nuigalway.ie, colm.oriordan@nuigalway.ie and pasi@disco.unimib.it} Abstract. Personalized Web search oﬀers a promising solution to the task of user-tailored information-seeking, and particularly in cases where the same query may represent diverse information needs. A signiﬁcant component of any Web search personalization model is the means with which to model a user’s interests and preferences to build what is termed as a user proﬁle. This work explores the use of the Twitter microblog network as a source of user proﬁle construction for Web search person- alization. We propose a statistical language modeling approach taking into account various features of a user’s Twitter network. The richness of the Web search personalization model leads to signiﬁcant performance improvements in retrieval accuracy. Furthermore, the model is extended to include a similarity measure which further improves search engine performance. 1 Introduction and Related Work Search engine users have diverse information needs, and it often happens that diﬀerent users expect diﬀerent answers to the same query [6]. In fact, given the potential of the same query to be representative of diﬀerent information needs behind it, personalized Web search has emerged as a promising solution to better identify the intended information need The usual approach to the personalization process in Web search involves incorporating user’s preferences into the retrieval method of the search system thereby moving from a “one size ﬁts all” approach to the customization of search results for people with diﬀerent information interests and goals. A signiﬁcant research challenge in Web search personalization is to learn about a user’s interests and preferences to build what is termed as a user proﬁle. The user proﬁle is the most essential resource within the retrieval model of a personalized search system. One of the main features that can be used to dif- ferentiate between existing solutions to Web search personalization is the source used when building the user proﬁle. Several kinds of sources have been explored by researchers in order to build a user proﬁle, with the most popular being search