The Realization Of Statistic Software Of Typed Error Rate In Chinese Text

In order to change the backwardness of the methods used in proof-reading in Chinese text and to improve the management and informatization construction in the press and publication industry, the General Administration of Press and Publication of the P.R.C established the scientific research item "Statistic software applied in editing and proof-reading error rates", attempting to develop a universal and standard software to test the quality of book editing and proof-reading in Chinese text. The paper, derived from the item, focuses on the technique research of the core techniques needed in developing and the function realization.The paper describes the techniques of automatic word-dividing and proof-reading, and analyzes the advantages and disadvantages of different techniques. Moreover, the paper proposes the improved technique modules for automatic proof-reading, which is based on the statistics of language model. Most importantly, the paper introduces the realization of the function applied in error rate statistics.Based on the related state standard of text editing and proof-reading quality, the taxonomy of editing and proof-reading in Chinese text, and the analysis of the rules in error statistics, the paper elaborates the design and realization of the software applied in error statistics in Chinese text. The three methods provided in the paper, that is, the method of classifying editing and proof-reading in Chinese text, the accurate statistic method of calculating errors and words, are applied in software for the first time.