Research And Implementation Of Mobile Reading Recommender Based On Hybrid Recommendation

With the advent of the Internet era, people’s lives have been greatly improved. People gradually entered the time of information overload from the era of lacking information. The Internet main access way is transferred to the mobile phone, which was traditionally dominated by computer terminal. So people can get information anytime and anywhere through the mobile phone. With the accelerating of life pace and increasing fragmentation time, more and more people use fragmented time to read. But how to select the books they are interested in and how to make the books to be found by other is a serious problem because of facing massive books. However, the development and wide application of the personalized recommendation system is a good way to solve this problem.Currently, mobile reading application is developing rapidly, and most applications are based on the traditional list-style-type or classification structure, and without function of recommendation. At the same time, most applications with recommendation function are based on association rules. According to the idea and method of software engineering, this paper design and implement the mobile reading recommendation system based on hybrid recommendation proceed from the phone user needs. The main works are follows:①From the background and significance of the system development point, understanding the development of recommendation system and the situation of home and abroad. And shortcomings of existing reading applications are pointed out.②This section analyzed the detailed meanings of the recommendation system and the mature developed recommendation methods are analyzed as well, such as content-based recommendation, collaborative filter recommendation, demographic recommendation and hybrid recommendation.③The fundamental requirements of the users were the basis of this research. The system functional requirements were carried out, and the corresponding use case diagrams and use case descriptions were presented. And then a specific design had been done on structure and processes for mobile reading recommendation system. And further detailed design of the system data platform was showed.④This paper designs the three modules of content recommendation engine in detail. From the point of mining user behaviors, this paper gets user interest model and implicit ratings, mixes the content-based recommendation and slope one recommendation. After filtering and ranking to the result, the final recommendation results are made.⑤Using Java to implement the system and displaying the interfaces of system server and client separately. Then, the system is tested systematically and comprehensively.