Employing latent dirichlet allocation for organizational risk identification

Any risk management method begins with risk identification intended to detect and classify potential risk items. This paper proposes a methodological process for risk identification by analyzing documents from all departments of an organization ignoring they are hard copies or electronic files. The LDA is used for texts classification which is the core work of the process. To do the text classification, Chinese natural language processing is analyzed. A review is also given on the existing probabilistic topic models. The results demonstrate the advantages of the proposed process and efficiency of the LDA based text classification for risk identification. It shows that proposed method not only improve the coverage but also boost the accuracy of risk identification.