Improvement of Learner Corpus and Development of an Online-Assisted Learning System Using Natural Corpus and Leaner Corpus

This project extends the research results of the applicant’s past NSC projects, in both 2007 and 2008, to improve learner corpus and then to use natural corpora and leaner corpus to develop an online-assisted learning system. On the basis of 1,563 articles established in learner corpus in the project in 2006-2007, significant research effort will be paid to correct the one of the factors of redacting the accuracy of collocation, MeCab segmentation errors, thus ensuring the accuracy of collocation. This project intends to combine the learner corpus and natural corpus, together with continuous collocation researches of the noun, the verb, the adjective, for developing the online-assisted learning system, in order to proceed with teaching studies. The research program is divided into the following three steps. The first step is to correct the half of the collocation of MeCab segmentation errors. Once the work is finished, we will recount high-frequency words in leaner corpus, compare them with the natural corpus, and use the results as the basis for follow-up operations. The second one is to establish the table of Japanese collocation, which is the necessary foundation of the online-assisted learning system. The last one is to develop online-assisted learning system. This project not only has close connection to the applicant’s previous researches regarding the Corpus Linguistics, but also enlarges corpus applications, thereby facilitating the development of Corpus Linguistics and Japanese language teaching in the future. In addition, the project will pay much attention to develop the online assisted learning system, and construct the multidiscipline research model of linguistics and information technology. The results of this project can be expected to make significant contributions to Japanese teaching and corpus linguistics.