A new automatic knowledge extraction method for course documents applied in the web-based teaching system

With the development of web-based teaching system, automatic knowledge extraction from course documents becomes more and more important. This paper gives a new automatic knowledge extraction method for course documents. Based on TF strategy, this method uses frequency and location to measure the credit value of knowledge. Moreover, the penalty factor is defined to adjust the credit value of knowledge. In this paper, the naive Bayes method is improved and applied in automatic knowledge extraction from course documents. Finally, we compare the method proposed in this paper with improved naive Bayes method based on experiments results. The results show that the average performance of this method is better than that of the naive Bayes method.