Material Select And Labeling Principle For Reading Material Corpus

This thesis can be divided into five parts.The first chapter is Introduction. It briefly introduces the key reason to build the Chinese Material Corpus and the referred relevant documents. It is essential to set up the Chinese Material Corpus due to fast development of Chinese language Teaching text books and the big influence of material selection for compiling a text book.The second chapter is the material selection's principle for the Corpus. In this chapter, 5 material selection principles are described: integrality, moderateness, practicability, interestingness and no timeliness. Among these 5 principles, integrality, moderateness and practicability are main standards for Chinese material selection, while interestingness and no timeliness are accessorial rules for reference.The third chapter is the labeling principle. There are 3 labeling principles described in this chapter: normalization, distinctiveness and Systematicness. Systematicness is the key principle for consideration while labeling, and normalization and distinctiveness should be thought about when setting up the labeling system.The fourth chapter reviews some of the works the author has done. In this chapter, the labeling of material corpus consists of two parts: artificial labeling and computer labeling. The property of artificial labeling includes: headings, authors, authors' nationality, abstracts, key words, character types, literature types, subject matters, nationality adaption, age adaption, adaption for ethnic Chinese and interestingness. Computer labeling consist of headings, authors, authors' nationality, abstracts, average sentence length, text length, new words density, difficulty, new words recurrence rate and words beyond compendium.The final chapter is summary and expectation. Author summarized the research achievement and deficiency, while looked into the future works. This paper set up a News Reading Corpus over 70000 words for Chinese language teaching based on research method of statistics and comparison, and author also completed material selection and labeling analysis. This paper provides a large amount of data and material for the compiling reference of Chinese teaching text books in the future.