Research And Development Of Recommendation System Hybrid Filtering

With the global development of informationization process and the popularity ofthe Internet, Internet has become one of the main ways to get new information resourceson the Internet gradually, with the explosive growth of, the user to extract the valuableinformation more difficult, some not to be concerned about the user information easilybecame isolated information,At present, for use in electronic commerce recommendation system mostrecommended effect is poor, low efficiency, high intelligent degree, already could notsatisfy the modern personalized needs, personalized recommendation system caneffectively solve these problems. This thesis introduces in detail the personalizedrecommendation related theory, elaborated the content based recommendation method,collaborative filtering recommendation method, data mining method and hybrid method.By combining the content based recommendation method and a collaborative filteringrecommendation method, uses two recommendation technology hybridrecommendation method. In a hybrid filtering technology research, summarizes thesparsity, cold start problem solution, the improvement of the existing classical mixedfilter algorithm is insufficient, put forward an improved algorithm and model.This thesis designs and implements a hybrid filtering method based on theelectronic commerce recommendation system, the system uses the B/S three layersystem structure model and J2EE architecture technology architecture. This thesisintroduces in detail the framework of J2EE, J2EE components and MVC design patternsof technical features, analysis of J2EE in three open source framework integrationtechnology of SSH and SSH integrated framework for the system realization process;used to J2EE concepts are introduced, and gives a set of J2EE developmentenvironment method and steps.This thesis is concerned with the system is based on a model of recommendationsystem engine model, it can provide the personalized recommendation, the new itemrecommendation, popular recommendation and other aspects of the recommendation,this thesis introduces in detail the recommended system design process, presents the core algorithm design, and came to the conclusion that.