On the semantic annotation of places in location-based social networks

摘要: In this paper, we develop a semantic annotation technique for location-based social networks to automatically annotate all places with category tags which are crucial prerequisite location search, recommendation services, or data cleaning. Our algorithm learns binary support vector machine (SVM) classifier each tag in the space multi-label classification. Based on check-in behavior of users, extract features from i) explicit patterns (EP) individual and ii) implicit relatedness (IR) among similar places. The extracted EP summarized check-ins at specific place. IR derived by building novel network related (NRP) where linked virtual edges. Upon NRP, determine probability place exploring Finally, conduct comprehensive experimental study based real dataset collected network, Whrrl. results demonstrate suitability our approach show strength taking both into account feature extraction.