A Framework for Identifying and Analyzing the Latent Risk Based on the Knowledge Discovery Techniques

摘要

In the global trend of competition becomes higher with every passing day, the risk threats also more easy to arise. Now, almost all transactions of the enterprise are recorded in the database. In database may be a risk which is latent such as the financial risk, operational risk, etc. The discovering the latent risk in the database is an important task of the risk management. In this paper, we propose an approach to identify and analyze the latent risk in the database based on developing and applying the knowledge discovery techniques. First, the authors builds a definition of latent risk and gives an illustration of the building of the function which identifies the latent risk, which is underlying for studying afterward. Next, we propose a framework for identifying and analyzing the latent risk in the database based on applying the knowledge discovery techniques. Then, a method to illustrate for the framework is introduced. The authors also build a strength level measure of clusters for integrating into a clustering algorithm to support for building the method. We choose objects are the customers in the enterprise and choose the type of risk is the receivable debt payment risk (a form of financial risk) and choose the strength level measure of clusters is the risk level measure to experiment. The experiment and evaluation have been performed to a real data set in an enterprise. The results obtained by the proposed method based on the framework will assist the managers in risk management usefully.