Research On High Quality Recommendation Reason Automatic Mining

High quality recommendation reason is that the research on automatic mining is a service recommendation system technology theory which is proposed on the basis of user-centered,in the context of overload data information and under the condition of users participate less as far as possible.High quality recommendation reason is that automatic mining system,which is on the basis of users' interested data information to analysis,then matching the users preference for the products,conducts an autonomous and individualized recommendation service and promotes user satisfaction.However,on vast information today,in order to solve the traditional recommendation system for the dark information mining is unknown and for the users interested in matching data information is not clear,the article adds to machine learning mining technology and grasping of users interested dynamic change,mines the user the new direction of interest constantly according to association rule technology and promotes the high quality and precise recommendation.Combined with the studying both at home and abroad,to improve the accuracy of recommendation system,the article mainly does further research from the following points:Firstly,introducing the traditional high quality recommendation system's research background at present,implemented contents and recommendation of the related theory of technology in detail,and the article does further research on the content which is on the basis of traditional recommendation system.Secondly,the article is on the basis of combination of recommendation algorithmic to combine,uses association rules of data mining technology to mine users' interested products,uses classification rules to classify the users and predicts users' interestingness according to the precision and recall rates,to maximum extent realizing the accuracy of recommendation.Thirdly,on the user modeling,to collect user behavior records as the basis,to analyze the influence which the recommendation attribute factors and influence factors have on the accuracy and high quality recommendation,to promote ‘certain people' information collection and ‘normal people' information particularity.Fourthly,according to accurate feedback of user interest and accuracy of user interestdata information collection,updating the user interest model timely,mining the potential direction of user interest.To improve users' satisfaction on the recommended products.By the MATLAB experimental results,the article indicates high quality recommendation reason is that automatic mining system has a good effect on the illustrated problems above.