Resumen de Study on Interactive Relations between Enterprise Social Media and DecisionStyle Based on a vector Autoregressive Model

Jinrong Mao, Mohammed Alhamam

• in the network technology increasingly innovation, weibo social media such as WeChat graduallybecome people thought together and important platform of communication, both theoretical knowledge and practical case studies, have proved that social media can be intuitive showpeople thefocus of the current problems, to provide effective basis for market development and investmentdecisions. Especially in the comprehensive promotion of text mining technology, topic extractionrelated to corporate social media has been reasonably applied in the financial market. However, due tothe problems of sparse and unstructured data, topic extraction in practical application has become themain subject explored by researchers and scholars. This paper studies enterprise microblogging social media, for example, using both the theme of the text content and platform interactive method, according to the relations between the two kinds of typical weibo relation and the matrix, and by usingtwo consecutive nonnegative matrix decomposition process is analyzed, finally the correspondingdistance as a result, and extracting keywords to express the relationship. The actual verification results show that the clustering accuracy and theme similarity of this method are more effective thantraditional theme extraction methods in corporate social media