Research on Approximate Web Page Identification Method in Online Social Network

In order to improve the user's Web experience and page retrieval performance, a new method based on key words and semantic matching for online social network is proposed. The first feature extraction of keywords from web pages, classify and merge of document concept, and establish the similarity model; similarity model based on feature vector is constructed, and the retrieval and storage of feature vector, identify the approximate pages in social networks. The experimental results show that the proposed method can effectively reduce the network noise, and has good recognition accuracy and recall rate.