Ontology Based Service Recommendation System for Social Network

The development of recommendation systems, such as traditional content-based, collaborative filtering and hybrid recommendation approaches have enabled the practical use of big data processing in WEB 3.0. In this paper, we propose an ontology based service recommendation system for social network. In this paper, implementation methods of the system are explained in detail. In order to extract user interests more exactly, the TF-IDF (term frequency-inverse document frequency) algorithm is improved according to the features of Micro logs and integrated with the Text Rank algorithm. Also, we have improved the Hownet based semantic similarity algorithm with the consideration of the density of sememe tree. Experimental results show that recommendation results of our system can well reflect the real interests of users, and the improved algorithms can make the results more accurate.