Research And Application On Information Recommendation Models

With the rapid development of the Internet, it has brought great changes to people’s lives. And people have entered an omniscient information era. Through the Internet, people can get a lot of information. However, with the overload information, Internet also brought some negative effects. Some traditional search engine technology have met people’s needs, But it need users take the initiative to search information. When a user attempts to use a search engine to search information, the traditional search engine technology will be returned to the user these pages including query keywords. The returned result pages may not only have many unrelated content, but also may lose some important key pages. How people can find the useful things in the ocean of information and realize intelligent recommendation has become a key study. Then information recommendation system arises at the historic moment. The purpose of the recommendation system is to build a bridge between information consumers and producers. So information consumers can find the useful information through information recommendation technology. In the meaning time, information producers can show the information to consumers by analyzing the data of consumers. Information recommendation is different from search engine technology. Search engine is a passive recommendation while Information recommendation is an active recommendation which needs study users’ characteristics. The recommendation algorithm is the core of the Information recommendation system.This paper first introduces the principle of the existing various recommendation algorithms, including based on demographic recommendation algorithm, based on content recommendation algorithm, collaborative filtering recommendation algorithm and hybrid recommendation algorithm, analyzing their advantages and disadvantages. Introducing the Slop One algorithm in detail, which is easy to implement and understand and has caused serious concern. It is also a kind of collaborative filtering algorithm. Slop One algorithm is implemented by calculating the difference between projects. The algorithm only considers the relationship between projects, so the recommendation’s accuracy is limited. In this paper, raise a user-based Slope One algorithm which improve the original weighted Slop One algorithm and take the similarity between users into account. In this paper carry out a contrast experiment between user-based Slope One algorithm and traditional Slope One algorithm on the MovieLens dataset.Experimental results show that the user-based Slope One algorithm is better in the mean absolute error against traditional Slope One algorithm. So the feasibility and validity of this algorithm are verified. In addition, in the end of the article, finish a simple movie recommendation system and realize the recommendation function.