Personalized Location Recommendation on Location-Based Social Networks by Efficiently Utilizing Spatio-Temporal Information

点击翻译按钮获取中文摘要

In location-based social networks (LBSN), people share location-related information with each other and leverage collaborative knowledge learned from user-generated and location-related content. Among various LBSN services, the location recommendation service suggests unvisited sites to the users based on the spatio-temporal information collected, which is a critical building block in many location-based services, such as the improvement of service quality, customer engagement and transportation management. The long-term objective of this research is the characterization of human movement properties from heterogeneous mobility data in order to understand the underlying dynamics of human mobility and to construct reliable predictive models for various spatial applications. The primary goal for the requested grant is the design and development of a set of locational prediction and recommendation methods based on human behaviour data on LBSNs and other relevant spatiotemporal datasets. The proposed research will make original and significant contributions in location-based services (LBS) and big spatiotemporal data mining research. New location recommendation models produced from this research will generate more intelligent and accurate location predication and recommendation. It will help built a new emerging paradigm of recommendation, which is to recommend “the right location at the right time” and provide the better understanding to the relationships between users' behavior related to external environmental, and users' internal activities that be reflected in their behaviors on LBSN. The research will have a wide range of applications, including local advertisement to promote the products and services based on users' locational preferences, intelligent personal and car navigation based on the user's intent, emergency response planning for community events, travel and tour planning, fleet management, geospatial gaming and many other location-based service applications. Mobility prediction enables appealing proactive experiences for location-aware services and offers essential intelligence to businesses and governments. This work will make a significant contribution to the development of this rapidly growing geospatial technology in Canada. Potential exists for the technology transfer of knowledge and software, with economic benefit to Canada. In summary, the proposal offers novel ideas, innovative approaches and distinctive solutions to location recommendation and its applications. The available research facilities, expertise and research experience provides unique opportunities to achieve the proposed objectives and train highly qualified personnel.