I.J. Information Technology and Computer Science, 2012, 10, 19-28 Published Online September 2012 in MECS (http://www.mecs-press.org/) DOI: 10.5815/ijitcs.2012.10.02 Copyright © 2012 MECS I.J. Information Technology and Computer Science, 2012, 10, 19-28 Location Based Recommendation for Mobile Users Using Language Model and Skyline Query Qiang Pu 1,2,3 1 School of Information Science and Technology, Chengdu University, Chengdu, China 2 Key Lab of Pattern Recognition & Intelligent Information Processing, Chengdu University, Chengdu, China Email: puqiang1116@gmail.com Ahmed Lbath 3 3 Joseph Fourier University of Grenoble, LIG-MRIM, Grenoble, France Email: ahmed.lbath@imag.fr Daqing He School of Information Sciences of University of Pittsburgh, Pittsburgh, USA Email: dah44@pitt.edu Abstract— Location based personalized recommendation has been introduced for the purpose of providing a mobile user with interesting information by distinguishing his preference and location. In most cases, mobile user usually does not provide all attributes of his preference or query. In extreme case, especially when mobile user is moving, he even does not provide any preference or query. Meanwhile, the recommendation system database also does not contain all attributes that can express what the user needs. In this paper, we design an effective location based recommendation system to provide the most possible interesting places to a user when he is moving, according to his implicit preference and physical moving location without the user‟s providing his preference or query explicitly. We proposed two circle concepts, physical position circle that represents spatial area around the user and virtual preference circle that is a non-spatial area related to user‟s interests. Those skyline query places in physical position circle which also match mobile user‟s implicit preference in virtual preference circle will be recommended. User‟s implicit preference will be estimated under language modeling framework according to user‟s historical visiting behaviors. Experiments show that our method is effective in recommending interesting places to mobile users. The main contribution of the paper comes from the combination of using skyline query and information retrieval to do an implicit location-based personalized recommendation without user‟s providing explicit preference or query. Index Terms— Location-Based Service, Mobile Information Recommendation, Language Model, Skyline Query, Implicit Preference I. Introduction Mobile-based applications have been grown as the development of mobile computing and communication technologies [1]. Since mobile devices are designed specifically for personal use, but having limited display sizes and processing power, we often need to select carefully the appropriate information to be presented to mobile users. Information recommendation for mobile users is currently an important research topic especially for location-based services using skyline query [2, 3] or spatial-temporal query [4]. Mobile information recommendation system is expected to provide more suitable and personal services to mobile user and overcome the shortages of mobile devices. Most of location based mobile recommendation systems take into account mobile user current location as well as his preference by using skyline query [2, 3]. Skyline query retrieves a set of interesting points from a large set of objects. For example, a hotel might be a skyline interesting for a tourist, if there is no other hotel which is nearer, cheaper than this one. The tourist can choose the most promising hotel from the skyline. Here the distance and price are the attributes of hotel which user cares for. So user‟s preference can be represented as a series of attributes of an object, that is also to say, users preference of interest will be affected by different dynamic attributes of the restaurant. As more attributes of an object are added into skyline query calculation, it will meet a time-consuming job [2]. Among all objects, if one or two objects are very strong compared to other objects, or when the number of category attributes is small, the resulting skyline may consist of a small number of objects [3]. Many researches focus on finding better algorithms to improve the effectiveness of skyline query,