Font Size: a A A

User Interest Based Mobile Client Video Recommend System

Posted on:2015-06-18Degree:MasterType:Thesis
Country:ChinaCandidate:H S FuFull Text:PDF
GTID:2298330467980548Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of smart phones, mobile internet, the information we contact is not only text message, but also more and more pictures, video and other rich multimedia information. To the users, the cost of select video is higher than the cost of select text information. To help users select video among the numerous video, and depends on user’s interests and preference to recommend the user favorite video becomes more and more important. This paper describes the design and implementation of user interest-based mobile client video recommendation system.This paper begins with the study of the characteristics of the video recommendation system, analyzes the necessity and feasibility of the development of video recommendation system. On this basis, this paper summarizes8important user requirements of the video recommendation system:offline playback, video recommends, video collections, etc. With the clear of those user requirements, this paper describes the details of the design and implementation of the video recommendation system. The core functionality for the video recommendation system is that the recommendation system can find user’s interest by user’s behavior such as watching a video, sharing a video with friends. The video recommendation system composed by the iPhone OS client and PHP based server. Among than, the iPhone OS client composed by video recommends module, user interest discovery module, video player module, automatic offline caching module. The PHP server composed by API module, retrieval module, web crawler module, offline behavior computation module, and memory module, etc.This video recommendation system using object-oriented design methods for design and development. Data store is using mongoDB, the NoSQL technology, who is popular in recent years. And the ElasticSearch, Mantle and many other open source tools are used to simplify the development of the system. BM25algorithm is used to calculate the recommended video score. With a rigorous sofeware testing, our system is stability and can meet users’needs.
Keywords/Search Tags:recommend system, Video recommend, iPhone OS
PDF Full Text Request
Related items