Keyword extraction for mining meaningful learning-contents on the Web using Wikipedia

The purpose of this paper is to provide a solution of extracting appropriate keywords to identify meaningful learning-contents on the Web. There are some issues in identifying documents that have learning content. Firstly, the documents need to be identified according to the learning area of a student´s school year. Secondly, the documents need to be identified according to the learning area that the student is now studying or studied. In this paper, we present a method of extracting keywords for mining meaningful learning-contents using Wikipedia. At first, we select the articles in Wikipedia with the arbitrary input keyword of learning items. Then, we select other Wikipedia´s articles related to the articles selected by the first process, using links and categories of Wikipedia. Furthermore, we calculate degrees of association between the articles and the keywords using PF-IBF, and put the degree on each keyword. Finally, we screen the keywords using his/her curriculum guideline to adjust the keywords to the learning area of the student´s school year. In the next step, we are planning to develop a method of screening keywords according to each student´s ability, so that we can select more appropriate keywords for each student.