Collaborative Recommendation Technology And Its Applied Research In The Scientific Literature Personalized Recommendation System

As the rapid development of Internet technology, Internet resources are quickly increasing. Technical literature electronic resources as the important part of Internet resources also sharply grow. Under the background recommendation system emerges as the times require, which recommends items to users according to users' interest. Recommendation system is also called personalized recommendation system. Collaborative recommendation is becoming a popular technique for reducing information overload. The main technique of collaborative recommendation is to recommend items to the active user referring to other users whose interest is similar with the active user.In this dissertation, we present two kinds of collaborative recommendation algorithms based upon the former collaborative recommendation methods. One algorithm is based on user and it recommends items to the active user by users' interest character. In the algorithm it acquires users' preference vectors referring to ontology concept vocabulary. The other algorithm is based on model and it makes recommendations by weighted association rules mining. It chooses the time of bibliography to weigh the resources.In the end we design a personalized recommendation system of technical literature electronic resources called PIRS (Personalized Information Recommendation System). The system is used to recommend electronic resources in the domain of avion. According to the algorithms the system can make recommendations not only by users' interest character but also by weighted association rules.