The Design And Implementation Of Recommadation System Based On Personalized Recommendation Engine Combination

In recent years, with the rapid development of the Internet, we has entered into the dataoverload stage. There are two solutions,one is information directories, such aswww.hao123.com; The other is search engine, such as Google and Baidu. Both have acommon characteristic that users clearly know what their needs; Unfortunately, The fact is onthe contrary.At this time, the recommendation engine came into being, it took the initiative torecommend to the user the information needed. it does not require user to provide any inputbut record the user’s behavior and conduct intelligence analysis in the background,and pushthe analysis results as personalized recommendation to the user;This paper introduces the research background, domestic and foreign research status andlevel of system-related theoretical foundation to explore the technology and tools of therecommendation system, and to determine the architecture of the system. Secondly, the detailsof the recommended system architecture, as well as the specific recommendation systemimplementation process. Finally, the the Mahout framework of the recommendation engine isused in large-scale data environments.The main contribution of the paper include:1. Respectively, realyze the three Recommendation Engines and generates arecommended list by the combination of the various recommendation strategies.2. Record and analyze user feedback to adjust the various recommendation enginecombinations; recommended items via personalized recommendation enginecombination; Experiments show that personalization engine combinationrecommended results better than using a single recommendation engines.3. Apply the recommendation engine to large-scale data environments by Mahout.