Research On Commercial Customer Credibility Analysis Method

In the 21st century, along with the science and technology development, and the commercial development of the credit question has caused everybody's attention, become common concern's issues of people. And feed industry management system accumulated largely customer data at the beginning, just do some simple query, without adequate use these data, and these data didn't play a role, but now with the businessmen realized they want to know more information, in order to get the results they want, companies realize the importance of these data, so, how to extraction, discrimination analysis from these sales data, it has become the right now it is important to solve the problem.Due to the sales data without certain regularity and complex, and usually data is largely, below this kind of circumstance, it is difficult to find useful information for the enterprise managers. But considering the factors of effect customer creditworthiness are limited, so it is effective that using decision tree algorithm to solve this problem.Compared with other algorithms, the advantages of decision tree algorithm are:1) Produce the classification rules simple to understand;2) The computing speed of decision tree algorithm is faster, at the same time in the accuracy is higher;3) Compare with Neural network or other classification algorithm, the decision tree can generate some rules allow users to facilitate understanding;Moreover, because the decision tree algorithm adopts the principle of plenty of examples, information on the characteristics of the information analysis, in this process, the calculation of each feature information gains, also find the information of reaction category and it is for this reason that just caught the question the essence, making this kind of method is very effective. And, another advantage of decision tree algorithm is more apparent when deal large amounts of data processing. Compared with other classification algorithm, decision tree algorithm's precision equals their precision, or sometimes has higher precision. So this paper chooses the decision tree algorithm analysis the feed industry management systems.With the increasing number of clients'transaction, the businessman accumulated the massive transaction data. In the past these data is only used for checking accounts, simple query and print for vouchers, data contains various hidden information not getting the attention they deserve, or due to technical limitations, it is hardly launched analysis. In recent years, with data mining technology use more and more, many businesses have begun to realize that how to use these data, get the useful information from these data. Personal credibility data mining thus becomes very active application fields.This topic studies data mining technology in personal credibility on the application of developing, using decision tree algorithm, provides the data processing method and potentially more information for enterprise customer credibility analysis, so as to facilitate the business adopt different credit scheme for customer.