Design And Implementation Of Text Information Recommendation System Based On Short Text Processing Algorithm Optimization

"Mai Quan" as the recommended polymerization product,base on "micro-blog"user behavior data including issuing,sharing,forwarding data to construct the user interest model,and rely on the model recommended information content to the directional user.how to recommend the information users accurate has become a new research direction,Therefore,the construction of user interest model is the most important link in the recommendation system,and the massive text information contained in micro-blog provides data analysis of user interest model for personalized recommendation.However,due to the shorter length of micro-blog text data,the user's interest diversification,the user is not recommended ideal results.Vector space model is the basic model of text structure representation,user interest model construction and similarity calculation.The text recommendation system based on short text designed in this paper,firstly,constructing a user interest model based on micro-blog user behavior data,Secondly clustering analysis the recommended text information and similarity calculation with the interest model,the results will be recommended to the user to browse,while the score data of user feedback recommendation system for users interest model updating.In this paper,the text recommendation system is based on the optimization of short text processing algorithm and the construction of user interest model.According to micro-blog short text data,in short text preprocessing,in order to enhance the ability of text feature representation,based on the weighted lexical entry in TF-IDF is proposed based on the semantic related short text feature words combined extraction algorithm.Then,in a short text clustering algorithm,due to the K-means algorithm sensitive to noise and dependent on the initial center selection,the paper the propose ideas of similarity measure is computed as the initial center,and the realization of the design scheme of the clustering algorithm is similar to the center of the cK-means passage based on simultaneous optimization algorithm based on short text the construction of user interest model.Finally,this paper applies the optimized user interest model to the specific text recommendation system,giving the overall framework of the system,describes in detail the implementation details of the core modules of the recommendation system and system test and analysis.The experimental results of the optimization of short text processing show that based on the semantic related text feature extraction algorithm and cK-means algorithm for text clustering accuracy is improved.Display on the transverse and longitudinal test show that text recommendation system based on user interest model,after the application of optimization of short text processing to the text of recommendation system,the accuracy and stability of recommendation systems have improved compared to before.