Design And Implementation Of Dating Recommendation System Based On User Interest Topic Model

With the fast development of internet?Browsing pictures shared on the social media becomes an important approach of making new friends who have the same interest.In the face of vast amount of information,more and more people hope to find their interested pictures and expect to follow the author sharing those pictures.How to rapidly acquire valuable interested user information has become the crucial issue.In recent years,related research on personalized pictures recommendation system has received more and more attention and become an important research field.Based on the research and analysis of the content-based recommendation methods at home,this thesis conducted researches on the application of recommendation system in the process of Internet dating and proposes a topic model based on user interest.The main work of this thesis includes following aspects.(1)Analyzed personalized recommendation technology and research status,educed the issue of image content representation and user interests' mining in the personalized recommendation methods.(2)Proposed a user image content and visual topic characters based personalized dating recommendation scheme.The scheme extracts the underlying visual features,uses the Bo F vocabulary model to describe the middle-level semantic features of the image,and uses the LDA theme model to describe the high-level semantic features.Based on the user's theme preferences of the image and user behavior,the hierarchical structure is established.The user interest topic model;the personalized friend recommendation method can well express the overall interest image feature of each user,the user interest model is more in line with the user's behavior law,and the recommended similar interest users are more accurate.(3)The personalized dating recommendation system based on the above scheme was designed and implemented,the function of the recommendation system was verified,and the performance of the recommendation system was verified by offline experiments on the image dataset.It was proved that the scheme is feasible and effective for dating recommendations.