Chinese-English spoken language translation question scoring method and system

Abstract

The invention discloses a Chinese-English spoken language translation question scoring method and system. The method comprises the following steps: preprocessing an obtained voice signal to be scored and extracting voice feature parameters; recognizing the voice feature parameters according to a pre-trained voice recognition model to obtain a voice recognition result; based on a preset scoring mechanism, scoring the voice recognition result to obtain a keyword score, a sentence meaning score, a grammar score and a fluency score; and inputting the keyword score, the sentence meaning score, the grammar score and the fluency score into a pre-trained score fusion model to obtain a comprehensive score of Chinese-English spoken language translation. The embodiment of the invention further discloses a Chinese-English spoken language translation question scoring system. By adopting the embodiment of the invention, the characteristics of the keyword, sentence carelessness, grammar and spoken language fluency of the to-be-scored voice signal are extracted and automatically scored, so that the scoring efficiency and the accuracy of the scoring result are improved, and time and labor are saved.

Description

technical field [0001] The invention relates to the field of computer-aided education, in particular to a scoring method and system for Chinese-English oral translation questions. Background technique [0002] With the development of computer science and technology, information technology has been widely used in modern education. It provides rich information resources and advanced teaching methods for modern education, making education forms, educational content, teaching methods, educational concepts and teaching environment There has been a change. On the other hand, with the development of artificial intelligence, acoustics, and linguistics, intelligent speech technology, as a new type of information technology, has gradually matured and been gradually applied to computer-aided teaching. [0003] The methods used in traditional speech recognition systems are mainly GMM-HMM (Gaussian Mixture Model-Hidden Markov Model, Gaussian Mixture Model-Hidden Markov Model) and natura...

AI Technical Summary

Problems solved by technology

However, in terms of grading, only the grading of objective questions is supported, and the grading of subjective questions still requires a lot of manpower and material resources, such as the Chinese-English oral translation question in the subjective questions of the English test. Therefore, in reality, for the oral expression level scoring task, Still completely manual review

Moreover, this kind of scoring task is not only highly subjective, but also time-consuming and intensive, so the quality of marking is difficult to control

Images