Development of a web-based concordance search system based on a corpus of English papers written by Japanese university students

This paper presents the process of developing a web-based concordance search system for supporting university students to write papers in English, their target language. The development of this system was based on a preliminary study of lexical analysis of a corpus of English graduation papers written by Japanese university students. The preliminary study is presented including a description of the data accumulated for this study consisting of approximately 900,000 words taken from 212 student graduation papers submitted over the last five years. The data compiled was divided into 10 sub-corpora by discipline, and further into four sub-corpora by rhetorical structure. 98.6% of the words in the AWL were found in the corpus, along with the 998 basic words in BWL 1, and 922 out of 988 in BWL 2, suggesting the corpus is suitable for concordance work with future students. A concordance system was developed on a self-study network accessible on and off campus. It was designed so that the users can study the collocation information for each search word or phrase in a cross-sectional way by discipline and/or rhetorical structure. The effectiveness of the system will be examined based on actual use by students after further development.