A New Annotation Method for Phonetic-Grammatical Study of Text Resources in Chinese Minority Languages

Abstract

An in-depth study of minority languages in China is not only beneficial to reveal the developmental characteristics of languages at a deeper level, but also to the development of related disciplines. Text resource annotation is a basic work and a work with relatively high manual involvement. How to ensure the quality of processing? How to improve the efficiency of processing? A range of issues such as these are the main reasons for the slow development of text resources. In the face of the study of minority languages in China, the corpus is being paid more and more attention, and thus the text annotation work is extremely important. This paper designs a set of effective solutions and methods to improve text annotation, which provides linguists with new methods for efficient speech transcription annotation and grammatical analysis annotation in text resource research, and provides a series of solutions for building large corpus, multilingual speech dictionaries, "interlinear cross-referenced" processed corpus, dynamic word lists, etc., with the dynamic and interactive nature between speech, text, dictionaries and word lists. Based on the annotation methods in this paper, hundreds of texts in Tibetan, Yi, Yao and other minority languages have been annotated and processed by several researchers, obtaining relatively satisfactory processing quality and efficiency, achieving efficiency and practicality.