Search

With the unceasing improvement of data mining and natural language processing technology, more and more researchers devote themselves to comment resources on the Web. While there are no readymade emotional corpora in the financial and securities domain, emotional analysis applied in this field is still rare. As financial information is usually in the form of unstructured Web Text, this paper fully considers characteristics of financial information, and analyzes the emotion of Web Text by calculating their emotional inclination values based on evaluation scores of morpheme. For each document we compute an emotional value. Its symbol indicates the emotional inclination, and its absolute value reflects the emotional intensity. Thus, this can prevent the limitation of lacking emotional corpus in the financial and securities domain. Experimental results demonstrate that this presented scheme optimizes the existing main schemes to effectively analyze the emotional inclinations of Web financial information.

As the rapid development of Internet technology, the influence to the financial fields of the financial information on the Web has become increasingly can not be ignored. Facing mass unstructured Web text about financial information, the paper present a method to calculate the tendency of text, and analysis the relationship between the emotional tendencies of Web financial information and the financial crisis of listed companies qualitatively. Via empirical analysis, Financial Warning for listed companies by using of Web financial information is not only feasible, but also ahead of financial indicators more than 3-4 quarter.