Abstract:

Knowledge context is the specific environment and background of knowledge creation and applications. Personalized knowledge recommendation systems integrated with knowledge contexts are also an important method to improve the reusing efficiency and sharing property of knowledge. In this paper we construct a personalized knowledge recommendation system based on context similarity. We construct a knowledge context model of multiple layers and multiple dimensions, and add knowledge contexts to the personalized knowledge recommendation system. We then calculate the similarity degree of knowledge context models. If the similarity of the current context reaches a specific value, the knowledge which is also relative to historical contexts, is recommended to target users. Experimental results show that the proposed method can improve the efficiency of personalized knowledge recommendation.

Key words: knowledge context;context modeling;similarity;personalized recommendation