Research On The Collaborative Filtering Algorithm Based On The Content Clustering

With the modern distance education platform emerging and developing, the number of teaching resources become digital and networking, and expansion of information resources. As a new education model , modern distance education platform In order to meet the basic learning needs of students functionality, network education platform system designing should be regard students studying networking service as starting point and ultimate aim ,which not only satisfy the based study needing of students but also can determine network resources and education platform services ,according to the type of professional students, demand characteristics, and other information, thus realizing the Intelligent Knowledge Network. Personalized Recommendation system to solve this problem is one of the ways.Recommendation system is derived from Information Filtering Technology. Information Filtering Technology can solve"information overload"and"resources wandering"preferably and make people more fully use of Internet resources. Information Filtering Technology is the foundation of personal information resources recommendation realizing. Based on the recommendation of information filtering technology has favorable prospects for the development and application. At present almost all large-scale commercial systems use different forms of Recommendation System. However, in the modern distance education network teaching platform recommended the use of the system is rare.In this paper, the main techniques of the recommendation system are deeply researched, analyzed and discussed. And then Research on the Collaborative Filtering Algorithm Based on the Content Clustering is brought forward. Based on The Distance Education Network platform of Shandong Normal University constructing, a personal recommendation system is designed and realized, which is imported in distance education platform and can provide customizing learning resources and items for students. The main work as follows:1,Deeply research the existing technology of recommendation systemFirstly this paper discusses the key technology of recommendation system, and then focus on the Collaborative Filtering Algorithm and the Content Filtering Algorithm and indicate them advantage and disadvantage, contraposing which find a better combination of methods and systems implement between CFA and CFA .2,Focus on Collaborative Filtering Algorithm and bring forward Collaborative Filtering Algorithm based on Nearest Neighbor ClusteringThe importance of nearest neighbor is considered and emphasized in Collaborative Filtering Algorithm based on Nearest Neighbor. Therefore, the similarities of users are ignored. If the number of training samples is great, the computation is huge. Then, the algorithm does not apply to a higher classification Text Classification, contraposing which the Collaborative Filtering Algorithm based on Nearest Neighbor Clustering is brought forward. CFANN has lower error rate, reduces the amount of classification to a large extent and accelerates the classification.3,Bring forward the Collaborative Filtering Algorithm Based on the Content ClusteringThe inherent limitations of Collaborative Filtering Algorithm based on Nearest Neighbor Clustering, which can not search all of nearest neighbors of aim and perhaps reduce the precision of recommendation. Therefore, this paper focuses on Ant Clustering Algorithm. A clustering method based on Changed Ant Combined Algorithm is presented, which can dynamic output the results of clustering without inputting some parameters such as clustering amount, initial clustering centers, the minimal distance between clustering centers, etc.. So, clustering will not be sensitive to such parameters. Thus, setting wrong parameters won't influence on clustering results. Finally, this paper discuss detailed and analyze this algorithm through experiments.4,The Collaborative Filtering Algorithm Based on the Content Clustering is applied for distance education platformA personal recommendation system is designed and realized, which can give students abilities to customize studying digital resource, and provide personalization service for students by using improved ant combined clustering algorithm combined with collaborative filtering. It mines transaction database of customized digital resources, and discovers mode of customized content, which are users'profile. System can give registered students personalization service and recommendation on the basis of association knowledge. It is a real-time commending system. So it can reflect students'changing interesting and needing. System captures students'visiting sequences real-time, and compartmentalizes it to a visiting mode. Also, with registered students'access history statistic, it can direct teachers and administrators what kind of resource to collect.5,Design and implement a distance education network platform which has personalized recommendation function.The paper introduces the design and implementation of personal recommendation system in Modern Distance Education Platform of Shandong Normal University in detail. The paper detailed discusses some crucial technologies of system implementation. Also the paper discusses the aspects of system expansibility and efficiency. Finally, the paper demonstrates the use of this system, and gives ideas of system improvement, anticipates the prospect of personalization service in distance education platform.