Fraud Detection in Tax Declaration Based on Bayesian Classifier

Taxing evaluation is that tax departments analyze tax and finance data of enterprises and find tax evasion enterprises. Fraud detection in tax declaration is an interesting problem. An approach of fraud detection in tax declaration based on a Bayesian classifier is proposed in this work. A Bayesian classifier is trained using financial data of sampled enterprises,the SVM is then employed to detect whether tax data declared by an enterprise is true or not. Experiment data sampled from typical 61 business enterprises of Qingdao city,Shandong province,China. Experimental results show that proposed approach is effective. The classification precision of proposed method is 93.55% in 31 sample data,and is 9.68% and 6.45% higher than that of See 5.0 and Support Vector Machine (SVM) based method respectively. Training the Bayesian classifier from 30 sample training data set takes only 0.698s.