Analysis System And Implementation Of Tcm Case Based On Text Similarity Score

In the text information processing, the calculation of text similarity has been widely used in the areas of Chinese-English automatic translation, information retrieval, automatic question answering system, automatic summarization, subjective question scoring and so on. It has been a hot and difficult issue in the Chinese text study for a long time. Traditional text similarity computing is mainly based on the frequency of same words and other information, while not to judge based on semantic information. This paper discusses all levels of text similarity calculation, proposes a new semantic-based text similarity calculation method according to the characteristics of the text, integrates this new semantic-based calculation method and the same-words-based calculation method, a integration-based text similarity calculation method is proposed, and the scoring accuracy can be increased. At the end, this method is applied to traditional Chinese medicine case analysis system. Specifically, the main work and results are as follows:1. Many calculation methods of words similarity are compared and analyzed. The HowNet-based words similarity calculation method is implemented.2. The main current text similarity calculation methods are analyzed, a new semantic-based text similarity calculation method is proposed and implemented.3. According to the text characteristics of this paper, a integrating same-words-based and semantic-based text similarity calculation method is proposed and implemented, and the effectiveness of this method is validated in this paper.4. The text similarity calculation method proposed in this paper is applied to traditional Chinese medicine case analysis system, the case analysis results are scored by the method proposed in this paper. At the end of this paper, this system is implemented by using C# language on.net development environment.The results show that the research of this topic and its results will have some reference value and application prospect for the text information processing.