Quality assurance for segmentation and tagging of Chinese novels in the Ming and Qing dynasties

This paper presents a word segmentation and named entity tagging project which annotates Chinese novels in the Ming and Qing dynasties. Computer-aided tools are used to assist the annotation. The focus of this paper will be on the quality assurance measures to ensure precision and consistency. The specification for word segmentation and named entity tagging is formulated based on the standards for modern Chinese segmentation commonly used in Mainland China and in Taiwan as well as the analysis of differences between Chinese classics and modern Chinese. The specification is established through iterative refinements. This refinement process can offer valuable insights into the quality control of computer-aided processing performed on Chinese literature works in the Ming and Qing dynasties and can be applied to those in even earlier periods. The finalized corpus, built in a computer-aided, manually-reviewed method in accordance with the specification, can be used for researches in literature, linguistics, information technology, and teaching of Chinese.