A Study On Personalized Information Recomendation Based On Weibo Tags

In recent years,with the quick development of the internet technology,especially the arrival of the web2.0,the internet becomes the information origins that connects the global,the resource on the internet has increased giantly.At the web2.0,social tags,the web information category method has been used by more web system,personalized information recommendation services have attracted attention via the massive growth of the information resource,while the traditional personalized information recommendation model can not comprehensively express users' various interests,also,can not reveal the resource features and the connection between resource theme and users.Although traditional personalized recommendation model take user-tags or user-resource feature into considered,but seldomly take users' social connection into considered,at the same time,users' comments data sparsity is also easy to lead to similar users discovery omit,thus leads to imperfectness of users cluster and failure of meeting users' need.Under this circumstance,this paper build an users interesting model based on the tags,the basis of users clustering is via adding user connection networks,use tags to represent users' interests,through users' mutual perpetual object excavates users potential connection web,find out the best interests match users to cluster.This paper has 6 chapters,they are:Chapter 1:Introduction,mainly introduce the research at home and abroad about tag-based personalized information recommendation model,includes main content?structure?problems to be solved and creativities.Chapter2:related technological theories,includes overview of social tags?personalized information recommendation theories and the shortages of tags personalized information recommendation theories.Chapter3:technologies about socal tags personalized information recommendation,includes users similarity calculation?users community found and users interests model build.Chapter 4:personalized information recommendation based on weibo label,includes the introduction of sina weibo?personalized information recommendation model build based on weibo?weibo users relationship minning,and finnaly construct whole algorithm framework model.Charpter 5:personalized information recommendation algorithm based on weibo label experiments and analysis,mainly take weibo data as sources,make the algorithm feasibility analysis,experiment design and verify the results.Chapter 6:summary and prospect,mainly summarizes the paper,prospects the future research work and puts forward the suggestions about the next steps research.