Research On Social Tag Recommendation Algorithm

Tag recommendation is an important research in social annotation systems, assist users in finding appropriate tags for resources and help consolidate annotations across all users and resources.But the inherent noise of the data and inconsistencies in the underlying tag vocabularies, have made the design of effective tag recommenders a challenge.In allusion to the existing tag recommendation algorithm shortcomings, in order to quickly produce accurate personalized labels recommend results, the main work of this thesis is:First,redundancy tag not only reduce the recommended quality of tag recommendation algorithm,but also impede the performance evaluation of tag recommendation algorithm,propose Subspace PSO Clustering algorithm to deal with redundant tag, after treatment,tag recommendation algorithm accuracy and recall rate has been markedly improved. Secondly,complement the advantages of the existing recommendation algorithm,apply weighted strategy and mixed strategy,then establish the recommender system architecture,on this basis,propose the linear weighted hybrid tag recommendation algorithm. We find our approach thinking is not only simple and easy to interpret,but also experimental results show that it has higher recommendation accuracy and computational efficiency.