Research On Warning Of The Product Harm Crisis Based On Surprising Modeling

In modern society, as the rapid development of world economy and information technology, the crisis, with higher frequency and growing destructive power, emerge in enterprises owing to the more unpredictable external environment, which brings more severe challenges for survival and development of enterprises. In recent years, under the influence of product harm crises, enterprises face not only sales fell or serious losses, but also consumer confidence decline, even bankruptcy. Enterprises are so frightened of product harm crises that more and more enterprises try to find the countermeasures by paying attention on how to judge the beginning and development trend of crises. It is imperative to research on warning for product harm crises.Based on surprising modeling, combined with the characteristics of the product harm crisis, this paper built a warning model of product harm crisis in order to form a warning monitoring method of product harm crisis and put it into practice. First, this paper established the method of product harm crisis surprising modeling. The study subjects of the method which based on cases that happened before are behavior and reaction of the reference group. Second, using the method, this study built a warning model of the product harm crisis. The details are presented as follows, data collection and processing, building network structure and setting association parameter. Data collection and processing is dealing with the cases through grading and inverted order. Through this way, the index can be determined and the cases can be instead by a series of information. The basic frame of the network structure is based on the stages of product harm crisis. So the building method is separating the cases with the stages and setting the network nodes with the reference group's factors. Then according to the relationship of the nodes, the arrows are determined. Based on the case information, association parameter which can be determined through statistics analysis is the quantitative relationship of the nodes in the network. Finally, using this model, the research helped an automobile enterprise to make a decision to handle an unexpected event through predicting the development direction of the event in the future. The model can confirm the most possible next step of the events in the crisis, and provide the best suggestion of the decision that can be made by the enterprise. So it can be easily used into the crisis management of enterprises, which helps enterprise to avoid serious consequences and decrease losses.