Research Of Collaborative Filtering Recommendation Algorithm Based On Discrete Multi-view Hashing

The rapid development of the information and internet technology brings us a explosive growth of resources and data.With the hope to improve the ability of information acquisition for users,recommendation system has attracted a lot of attention.Most of existing recommendation method based on collaborative filtering only concerns user information from single view.And it brings a bad influence on the efficiency of computation and storage because it makes computations between high-dimensional vectors in the process.In this thesis,we proposed a high-efficiency and high-quality collaborative filtering recommendation algorithm based on discrete multi-view hashing coding technology.The main contribution of this thesis are as follows:Firstly,we proposed a method for multi-view anchor graph construction based on multi-view locally anchor embedding to fuse data from different views.Our proposed method exploits both view-specific and inter-view information to make an effective fusion of multi-view data.Secondly,based on the proposed multi-view anchor graph construction method,we proposed a CF hashing algorithm for multi-view data to learn similarity-preserved binary hashing codes.We choose to encode both users and items to make a maximum usage of information from varied views.We also proposed a method to calculate codes for out-of-sample data.Finally,we proposed an effective and efficiency algorithm based on the proposed hashing learning model.We prove our algorithm through experiments on several real-world benchmarks.