Research And Implementation Of Activity Recommendation System Based On Douban Events

Merely online interaction through the social network have cannot meet the needs of users. What users need nowadays are the activity platform which strengthen social relationship by combining interests and activities together. Event-based Social Networks(EBSN) subvert users' social habits from traditional friend first and then understanding turn into the first cognitive then friends. As a new social network, EBSN set “interest” as the core point of social through online subject and offline activity to know each other first, make friends and then cognitive scope with cope within the margin of safety. This pattern can make more realistic and effective social relationships.No interest means not partner, no activity means not relationship. Interest is the raw power to driven communication, but activity is the direct method to strengthen relationship. For the modern urban people, both are indispensable. So the future of social is necessarily go on the road to social activities based on interest. This is also the reason that EBSN is so prevalent in current and become a new hot spot in a period of time in social networks. Therefore, we studied this kind of offline events based on EBSN by analyzing the event properties, the behaviors of attendance, and the influence between online and offline events. On this basis, we propose an activity recommendation from where user stand and activity organization recommendation from the view of host. The main work is as follows:1) Through the open APIs provided by Douban Open Platform and a crawler to realize the data collection of user, activity and social relationship information. Then describes statistical properties of the collected dataset, and analyzes the event properties, the pattern of user behaviors and the influence between online and offline events with the combination of questionnaires and experiments over dataset.2) After discovering a set of factors that connect the physical and cyber spaces and influence individual's attendance of activities in EBSNs. These factors, including content preference, context(spatial and temporal) and social influence, are extracted using different models and techniques. We further propose a novel Singular Value Decomposition with Multi-Factor Neighborhood(SVD-MFN) algorithm to predict activity attendance by integrating the discovered heterogeneous factors into a single framework, in which these factors are fused through a neighborhood set. Experiments based on real-world data from Douban Events demonstrate that the proposed SVD-MFN algorithm outperforms the state-of-the-art prediction methods.3) We investigate how to select potential participants based on the preferences and social influences in EBSNs from an event host's point of view. We formulate the problem as mining influential and preferable set in social networks, and study this problem from two complementary aspects. To solve this problem, we propose a novel Credit Distribution-User Influence Preference(CD-UIP) algorithm by combining user preference and influence maximization. The results of our experiments demonstrate that the proposed algorithm outperforms the state-of-the-art prediction methods.