Reaearch On The Technology Of The Personalization Recommendation In The Mobile Search

With the popularity of the3G network and the development of intelligent terminals, more and more people in the world get access to the internet via mobile phones. Compared with the traditional internet model, the mobile internet also has a very large demand for the search, because it is very convenient to search for the information anywhere and anytime in moving state. In the future, one of the main trends of China’s mobile internet is mobile search. Although mobile search can not substitute the traditional search now, the mobile search aimed at users for environmental factors will be popular. Under this background, this paper mainly researched personalization of the mobile search for mobile users.Along with the popularity of smart phones and its increasingly powerful function, the personalization of mobile terminal applications has been put forward. Users want to change their phones’interfaces, styles and layouts of phones according to their own preferences and habits. In addition, limitations of mobile terminals themselves also require that mobile terminal applications are more intelligent, such as small screen and keyboard.Now the mobile search, in different background, for different people, at different places, uses the same search engines and keywords, and the results may not be the same. On the one hand, the personalization refers to personalization of the search results. Even if different users search for the same words, the results may be not the same.This paper first introduced the development of the mobile search and personalized recommendation, and analyzed the related technologies about personalized recommendation, and then researched two different types of personalized recommendation technology and the development technology of Android applications. According to the characteristics of mobile terminals, a design scheme of a mobile search client system was proposed and the technical implementations of key modules were described in details. Then the typical personalized recommendation technologies were studied in-depth. On the basis of the traditional Slope One algorithm, this paper described an improved Slope One algorithm which brought in the conception of similarity between users. Finally, the design scheme was applied in the project "the mobile intelligent semantic search platform based on the theme" and experiments demonstrated the correctness and feasibility of the scheme.