The Study Of Reciprocal Recommendation Algorithm For The Field Of Online Dating

With the development of electronic commerce,more and more two-way recommendations between user and user have been appeared,the traditional personalized recommendation between project and user has been unable to meet the needs of users,which give birth to the online dating sites as the representative of the development of mutual interest.The online dating sites provide a great deal of dating information every day for the opposite sex,the job search website not only provides a lot of job information for job seekers every day,but also provides a lot of information about job hunting.However,these information is very complex and lack of effective system classification,whether it is for both friends or to participate in the recruitment of job seekers and recruiters,it is difficult to accurately locate the object of interest.Therefore,it is urgent to improve the recommendation algorithm and improve the recommendation quality to the greatest extent.In this paper,we propose a reciprocal recommendation algorithm for online dating,from the two aspects of matrix completion and reciprocity similarity to improve the efficiency of the reciprocal recommendation algorithm,we give full consideration of the two sides and then give further study for reciprocal recommendation,the main work and contributions are as follows:(1)Aiming at the problem of data sparsity in the reciprocal recommendation system represented by online dating,we propose a matrix completion algorithm to solve the problem of data sparsity.Firstly,we analyze two kinds of matrix completion methods deeply.Secondly,we combine the two methods with hybrid weighting,and then put forward a hybrid weighted matrix completion algorithm based on K-Means and LMaFit,the low rank matrix completion LMaFit algorithm and K-Means clustering algorithm are complementary.Lastly,the matrix completion algorithm based on weighted LMaFit and K-Means mixed is better than anyone alone in the mean absolute error of MAE through the experimental verification.(2)In this paper,a reciprocal recommendation algorithm based on reciprocity similarity is proposed.Firstly,the explicit preference and implicit preference of male and female users are defined,and give the computing method about explicit preference similarity and implicit preference similarity between male and female users.Secondly,assign different weighting factors for them to form the improved reciprocity similarity according to the effect of explicit preference similarity and implicit preference similarity between male and female users on the reciprocal recommendation.Lastly,compared with the two recommendation algorithms commonly used in the current recommendation,the proposed algorithm is significantly better than the other two algorithms in terms of accuracy,recall and harmonic mean.