The Research On The Identification Of Suspicious Financial Transaction Accounts Based On Data Mining Technology

The crime of money laundering is a public hazard, not only encourages the social crime, but also disturbs the financial order, and damage the image of our country. With the development of economic globalization and the improvement of the international financial market, anti money laundering has become an important issue in the field of international finance and others. Anti money laundering law enforcement is mainly dependent on the financing institution report to the court and government agencies, that is, large and suspicious transactions report. In China, there are problems such as large amount of data report system, for example:huge data, high false positive rate, easy to be avoid, uncertainty decision, low learning ability and so on. This paper is on the base of the current data reporting system, brings forward suspicious financial transaction identification method which base on decision tree, and also not easy to avoid, dynamic deciding, and self-learning. It has practical significance in a lot of fields, and also in line with the development direction of the global anti money laundering research. Decision tree is one of the most important research aspects in data mining. It analyzes huge complex and noisy data, although decision tree method is a very effective method for data mining in the identification of suspicious financial transactions in the anti money laundering, it is important to find out the key decision attributes, it changes the effectiveness of the ID3 algorithm even in the whole anti money laundering work.Based on this, first this paper brought a brief introduction to the development history and development trend of data mining technology, then made a comprehensive introduction in technical analysis of the point of view of the data mining mainly technology decision tree classification, and discussed the representative algorithm of ID3 decision tree algorithm, analyzed and pointed out the main advantages and disadvantages. Finally, we built a suspicious financial transaction risk assessment model, assessed the risk of money laundering rules set by mining analysis, explained the process of data preparation, data conversion, calculating information gain, generating decision tree and result display in detail based on real data., achieved a good classification effect after the test, and finally pointed out the further research direction in the future.