Research Of Personalized Internet Advertising Recommended System Based RSS

With the rapid development of the Internet, Internet Advertising has been rapid development. Compared with traditional media advertising, Internet Advertising through multi-media, all-weather(smy), global exposure, with low-cost, highly interactive features. Internet Advertising favored by more and more advertisers, and becomed the research subject of many scholars.In recent years, both in commercial application and research field, Internet Advertising has been a great development. However, with the explosion of Internet Advertising delivery, the click through rate(CTR) of the banner falling. The prospect of Internet advertising has been questioned. In order to increase the CTR of Internet Advertising, providing personalized information for each user has been a research focus in the ad industry and the academic currently.An algorithm for mining frequent patterns based on FS-tree (1 item frequent sub-tree). Online RSS readers have a large number of registered users, it leads to very large data sets. So if use FP-growth algorithm mining association rules for them, it need to occupy a lot of memory space. To solve this issue, the algorithm for mining frequent patterns based on FS-tree has been put forward. The algorithm enables the FP-tree the greatest depth and width to the largest separation, through constructing FS-tree. Experimental results show that the algorithm occupys small memory, the implementation of efficient and effective conduct of the association Mining Rules.An algorithm based item classification collaborative filtering. On the online RSS reader site, the number of feeds by users subscribed accounts for only a very small number. So the number of users and Feeds sharply increase will lead to score matrix become extremely sparse, resulting the very scarce items with two users of the common score. So it does not accurately calculate the neighbors similayity by collaborative filtering algorithms, and lead to recommended quality a sharp decline. Against the problem, it put forward the recommended algorithm based item classification collaborative filtering. The algorithm uses the results of Feeds classification through system, to calculate the similarity between the various items, and then through the method of weighted, to forecast the score of items which did not been score, thus effectively reducing the sparsity of data. The results showed that, the algorithms can effectively resolve the defects of the traditional similarity measurement methods, when the score in the sparse data, and significantly improve the recommend system's quality.The strategy of smart advertising recommended. An effective ad recommendation system should have the capacity that adjust the output of sets of ads recommendation, according to variety factors, and set ads filter rules for the demand of advertisers. So, the paper set a number of filtering rules, and put forward a mixed recommendation strategy. It can enable the system generate different sets of ads recommendation according to background of issues and the actual situation. It will call collaborative filtering algorithms to get sets of ads recommendation for registered users. Then registered users can get the personalized RSS ads when they login. With access to in-depth, when the length with current view sequence of users reach a value, it will trigger association rules recommended algorithm, otherwise trigger classification recommended algorithms to get sets of ads. The mixed ads recommendation strategy efficient solve the issues of new users and new projects. Both collaborative filtering algorithms and association rule mining algorithms are completed offline, so the system has better real-time.Based on the decision tree model of the loss of advertisers. The management of the loss of customers is many industries concerned about an important issue. Advertisers are the main source of income of online RSS reader, it should be great concern to the loss of them. To address this issue, based on RFM model of CRM, this paper put forward the models of the loss of advertisers, according to the actual situation in the advertising industry, and using decision tree algorithm to analsis the model.