Risk element identification in Bulk Stock Electronic based on the text of domain news

Chongyang Wang, Xiaoyan Gu, Xiangping Deng, Dengsheng Wu

Abstract: Risk elements are crucial to our study of commodity e-trading risk, and how to identify the elements of commodity e-trading risk is important to risk analysis. To solve this problem, based on the risk factor recognition method based on text mining technology, syntactic dependency analysis is integrated to analyze the sentence structure and semantic relationship, reduce the interference of meaningless content in the text to the clustering result, and improve the recognition effect of risk factors.so that the identification of risk elements is more comprehensive and objective. In the empirical analysis, the topic model in text mining was improved, based on dependent syntactic analysis techniques were integrated, and the effectiveness of the model was significantly improved when comparing before and after the improvement. The risk elements identified not only include almost all widely studied elements, but also some elements rarely mentioned in the existing literature, which illustrates the feasibility and effectiveness of the method. The method provides an important complement to traditional qualitative analysis methods, and the identified risk elements are an important reference in risk analysis, assessment and early warning of the industry.