Research On Abnormal Transaction Detection Of Banking Account

• Main Authors: Chi-Nan Chuang, 莊啟男
• Other Authors: Chih-Cheng Lien
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/11336130761882307050

碩士 === 東吳大學 === 資訊科學系 === 95 === Because the change of environmental competition of the finance and financial industry's E-commerce result, a lot of financial products and tools appear on the market. Bank customer can easy through various kinds of tools , he can draw out money or transfer the money to another account. The development of these is bringing a lot of facilities in life to us, but the same one offers the chances of crimes to a lot of ruffians too. The ruffian utilizes the illegal method to duplicate the cards and steal the password, or the hacker analyses and explains customer trading website's password in the bank, in order to fraud the money in customer's account. So we hope to use proper way, offer correct detecting information within time immediately. To warn customers or the banking system of bank by preventing the losses, and make the customer of the bank more secure. This research uses relation analysis and algorithms of decision tree to build and construct a decision tree and detecting rule for abnormal transactions. To utilize tools of data mining (WEKA) and verify that reject the attribute of zero coefficient, can simplify and accelerate the time for the decision tree analyse. We through this way to detect the abnormal transaction and create a detecting abnormal transaction model by a bank account.