Text categorization in selecting authentic materials on tertiary level

Corpus linguistics as a methodology of linguistics research has had such significant influence over the years that corpora have been used extensively in language teaching and learning. This paper explores, through the method of text categorization (TC), the selection of appropriate corpus data for tertiary-level students on English learning. It firstly introduced the importance of corpora data´s authentic nature on teaching English as a foreign language (TEFL). Then criteria for selecting authentic English materials on tertiary level were given. According to the criteria, it discussed about the method of TC and by which it conducted the experiment of classifying corpora data to meet tertiary-level learners´ needs. Results show that TC is an effective way of categorizing corpora data and thereby selecting appropriate authentic material considering specific criteria. Additionally, through TC, corpora data could be applied as valid teaching and learning resources in TEFL.