A Research About Social Search Rank Based On The Hidden Collaborative Model

After decades of development, the Internet has blended in the real life of the people togreat extent, with the development of the industry and the demand, the Internet is divided intoseveral large entrances, including search engine, browser, instant messaging andcontemporary popular social networks, and so on. Search engine aim to solve the problems ofpeople obtain valuable content in the massive Internet information quickly and easily. In thevirtual network, social network established an interpersonal network to narrow the distancebetween each other.As the two major Internet portals, search engine and social network cannot develop inisolation. Traditional search engine will return the same search results on the same search forany user’s request. During the transformation process of personalized service, search enginejust based on user’s interests factor to provide personalized service for user separately,consequently, user’s personalized service results cannot learn each other. Social networkprovided a good foundation platform for users to learn search results each other, in this way,during the search the user is no longer alone but synergy friends together to complete a searchtask. The integration between search engine and social network has given birth to the study ofsocial search.However, the study of the social search is still at the stage of the preliminary stage, as tothe problem how social search engines and social networks combined many studies maintaindifferent understandings. This paper study from the direction that the social network offeredcollaborative service to the search engine, on the other hand, it can be called social searchbased on hidden collaborative web search.The main research work includes the following aspects:Firstly, we aim to analysis search engine log in the way of social network analyzing, andrepresent logical the search engine’s logs and web links in uncertain graph way. In addition,based on uncertain graph SimRank algorithm, calculate the query words and the similarity ofthe page, and the final result is to establish page description libraries with weighted similarityand query words.Secondly, starting from the analysis of user search experience, calculate the trust of usersin the social network. On the basis of establishing trust measure between users proposed thesocial search sorting algorithm.Finally, after combination of the two aspects above, we comprehensive proposed social search sorting algorithm.