Automated Chinese-English translation scoring based on answer knowledge base

The current automated Chinese-English translation scoring methods mainly fall into two categories: one is based on the text similarity and the other on Natural Language Process (NLP). Both of them have to face the problem in their accuracy. This paper proposed an automated Chinese-English translation scoring method based on answer knowledge base by analyzing the shortcomings of the current automated scoring methods of Chinese-English translation. Answer knowledge base is constructed by means of such methods as semantic expansion of the answers. Students´ answers are automatically scored based on the answer knowledge base, which improves its adaptability, flexibility and accuracy. The analysis of contrast experiments indicates that the method can work well and its accuracy is obviously improved. Meanwhile, this method can also apply to the automated scoring in other subjective tests.