关键词: 路线搜索, 群体推荐, 签到数据, 基于位置的社交网络

Abstract: Location-based social networks have generated a mass of data that can reflect user’s preference and the regularity of popular routes. These data have provided a new mode in searching travel route. The existing research on group travel route search usually aggregates user’s preference and then performs the search by using individual user route recommendation algorithm. In real life, when group users browse a global route, individual user wants to select different attractions in the local area of attractions in the route. Based on this requirement, this paper proposes the problem of local distributed group travel route search with the goal of getting a travel route which has local distributed POI (point of interest) and can make high satisfaction for the overall group. The search finds the optimal route using transfer graph which generated by the check-in data. In order to improve the efficiency, this paper proposes a double-hierarchy transfer graph generated with the popularity and metastasis of POI, and achieves hierarchical query. Designing an optimization algorithm based on branch and bound search strategy further improves the searching efficiency by using the controlled relationship between nodes. Using check-in data sets from Gowalla and Foursquare social networking websites, this paper evaluates the proposed algorithms with extensive experiments on the route profit and searching efficiency, and verifies the effectiveness of the algorithms.

Key words: route search, group recommendation, check-in data, location-based social networks