Personalized Information Recommendation Based On User Profile

With the fast increasing of the amount of data in the internet,users face with “Information overload”.It's essential to help users quickly find the needed information from the massive data and to provide personalized service to them.As users get more and more data on the Internet every day,it has become a trend to give personalized the recommendation for the use of these data.Nowadays,the most popular social network--microblogging attracts hundreds of millions of microblogging users and the resulting data is huge.Mining user interest from the microblogging data and giving the personalized recommendation is very meaningful.Using of single data source and the simple model,the traditional micro-blog recommendation algorithm results the low recommendation accuracy.Therefore,the primary works described in this paper to overcome such issue include the following aspects.Firstly,a personalized user recommendation algorithm based on User Profile was proposed.By analyzing the significance and correlation of individual user data,such as text,label,social relationship,and personal information,the algorithm then generates new labels and suggests related interests with training LDA model and SVM classifier.The user's interests were assigned by weighted sum of these factors.The overall recommendation accuracy is improved.And user recommendation algorithm based on User Profile is better than the traditional VSM and HMM model in allowing users to have a better micro-blog experience.Secondly,a personalized recommendation algorithm based on User Profile is proposed.The similarity of User Profile and micro-blog Profile are calculated.According to the social characteristics of microblogging users,the method to calculate the intimacy between users is proposed with the behaviors of comments,reviews and collection.Then the microblogging score is calculated by combining the similarity and intimacy.And the scores are used to arising the quality of micro-blog recommendation.Comparing a recommendation algorithm based on User Profile with VSM-based microblogging personalized algorithm,the experiments show that the accuracy and recall ratios are higher.Thirdly,based the algorithm and combined with the actual application,a microblogging personalized information recommendation system is designed and implemented.The system is divided into four modules,crawling model?processing model?creating model and recommendation model.