An Implementation of ID3 Decision Tree Learning Algorithm for Tax Fraud Control and Prevention System
Oshoiribhor Emmanuel O, John-OtumuAdetokunbo M., Ojieabu Clement E

Every month business ventures pay certain amount of money as tax to the government agency in-charge of collecting tax as internally generated revenue. This tax amount ought to be a certain percentage of their earnings or profit, but the agency in-charge has no structured means of apportioning or predicting the amount of money to levy the tax payers thereby over-estimating or under-estimating the tax amount charged. This scenario has posed serious financial fraud issues of cash suppression and diversion in the current tax collection system. For the purpose of this research work, ID3 classification technique based on decision tree has been used to properly classify tax payers into tears in order to monitor, control and reduce fraudulent tax activities in the present tax collection system. The result of this research work will assist the state government to control tax fraud of different kinds, and to properly predict the expected tax income to improve on the developmental projects of the state.

Full Text: PDF     DOI: 10.15640/jcsit.v4n2a4