Research On The Construction And Application Of Book Recommendation Service System

With the popularity of network, the Internet era of information-overloaded has arrived.How to filter out the irrelevant and useless information, and directly display the informationthat users are interested, is a very important issue for the Internet development. In the field ofbooks on-line sales, Amazon.com and Dangdang.com have applied a variety of informationtechnologies to improve the book recommendation service. But the book recommendationservice system still has a lot of deficiencies, such as un-precise individual recommendation,un-intelligent book retrieval methods, lack of social recommendation etc. These problems areemergent to be solved.Therefore, in order to solve these problems, this article build the book recommendationservice system on the basis of innovative recommended approach, including three maincontents of individual recommendation, intelligent book retrieval method and socialrecommendation. They go as follows:Firstly, individual recommendation, including the joint recommending method based oncontents and users and the joint recommending method based on readers and writers. Theformer method subdivides books category and users to calculate the neighbor users instead ofusing the original calculating method. Then the system can recommend by collaborativefiltering method, including the same type of books, different types of books and the neighborusers. The latter method also use the same subdivision way to calculate the similar readersand authors by comparing their files, then it can recommend by collaborative filtering method.On the one hand, the system can recommend similar authors and his works to the reader. Onthe other hand it can recommend similar readers and his other types of book interest to theauthor, then the author can get more inspirations.Secondly, intelligent book retrieval method, including intelligent quick retrieval methodand the optimized category retrieval based on association rules. The former method accuratelyidentifies and analyzes the users need by big data mining techniques, then achievesaccurately and intelligently matching recommendations from the background mass ofinformation resources by the big data and cloud computing technologies. The informationresources come from the books repository,"the answer library" of "secondary index" and the artificial synergistic knowledge base named "know of books". The latter method firstlyproduct users browsing sets by their browsing records on category retrieval, then calculatethe frequent item sets, lastly calculate the association rules which meet the requirements.After all these, the association rules can be applied to optimize the category sort of retrieval.Thirdly, the social recommendation, including independent social network constructionand network recommended alliances. The former method has greatly improved the onlinebookstores social network features, which achieves a virtuous developing cycle of book salesand social network. The latter method has companies and individuals joined the networkrecommended alliances, in order to bring more customers into the book stores.Finally, considering the high degree of difficulties and too much works of implementingthe system, the article carried by way of questionnaire to analyze the feasibility of the system,which confirms the value and advantages of the system. And it turns out that the system ishighly accepted by the market.