Application Of Recommendation System In Distance Education

Nowadays, distance education is attracting more and more attentions. It can break the space-time limits, and provide the learning pattern and platform with independence, freedom and convenience, and realize the online education at anytime and anywhere. With the rapid development of internet and information technologies, distance education has transformed from traditional type to a modern pattern with character of digitalization, networking and informatization. However, due to the extension of distance education on quantity and scale,the education resource on the internet is exploding, so that users face the troubles of information overload when they enjoy the convenience of internet. It's difficult for users to pick up the information which is suitable for them from the mass resource, so the use ratio of resource is lowering down. Thus,the individualized service for the users is becoming a problem demanding prompt solution. Recommendation system is an effective method to analyze and provide information resources which conform to users' demand.This paper makes an introduction of the classical recommendation algorisms and demonstrates their principle, working process, advantages, disadvantages and application characteristics. On the basis of classic recommendation algorithms, this paper proposes characteristic distance based collaborative filtering strategy, CD-CF, which is aimed at improving the performance of traditional collaborative filtering when the user evaluation matrix is sparse. CD-CF imports the conception of characteristic distance, which can reflect the differences of users more intuitively. The strategy integrates the characteristic distance as another criterion into the similarity calculation according to processing hybrid method, and amends the classical similarity calculation process, so that to ensure the process of nearest neighbor set forming is more reasonable. The article describes the CD-CF in detail on its conception,principle,and demonstrates the CD-CF's features and applicability.To provide a solution on the information overload problem, this paper analyzes the data model and processing scheme of the distance education system, and integrates the CD-CF algorithm into the education platform. Also, this paper evaluates the strategy's performance by experiments with real distance education website's data, and compares the evaluation to other recommendation algorithms. The experiment results indicate that CD-CF strategy is efficient and accurate for the distance education recommendations. CD-CF has application prospect and applicability.The conclusion summarizes the research work in this paper, and makes a discussion of the problems and advantage of the application of recommendation system on distance education. The expectation of the research work is proposed at last.