Building Chinese Word Knowledge Base for Children's Leveled Reading

Zhiying Liu,Lijiao Yang,Jiaomei Zhou,Lu Zhang

DOI: https://doi.org/10.1109/IALP51396.2020.9310486

2020-01-01

Abstract:With the development of great Chinese education, the domestic leveled reading of Chinese has attracted more and more attention. Both schools and parents urgently need a reading system that meets the development of children's reading ability. The hierarchical construction of words as the carrier of reading materials is even more important. The difficulty level of words has a direct and significant impact on the text complexity of reading materials. This paper focuses on the construction of the Chinese Character-word grading of the Chinese reading system, and attempts to establish the Chinese characters knowledge base with Character ranks in line with the characteristics of Chinese characters themselves. In terms of the Chinese character knowledge base, this paper absorbs the research results of exegetical studies, and determines the hierarchical attributes of Chinese characters including shape, meaning, and word formation ability of Chinese characters, builds the Chinese character knowledge base for leveled reading containing 3350 Chinese characters with features. As for the word knowledge base, this paper describes the attributes of part of speech, word meaning, context, etc., especially the use of Hierarchical Network of Concepts theory to define the level of difficulty about the cognitive attributes of semantic categories, and finally builds a Chinese reading leveled word knowledge base containing 18300 words with features covering shape, meaning and context. Based on it, the content of words, the word density, the proportion of super-class words, the number of class symbols, IOG and other attributes are described to guide the automatic grading of Chinese texts which got a better result.