Using Machine Learning Technique to Detect SMEs Tax Frauds in Nigeria

ABSTRACT:

Tax fraud in Nigeria is on the increase and has led to the international community rating of Nigeria as one of the developing countries with a high rate of tax fraud; small and medium enterprises (SMEs) inclusive. Due to lack of introduction of advancements in technology to the SMEs taxation, government has lost most of the expected revenue from that source. Despite the efforts employed by the government by engaging tax agents such as Board of Internal Revenue Direct Taxation (BIRDT) and Npower tax, the fraudulent characteristics still persist. The aim of the study was to develop a model for the detection of tax fraud in Nigerian’s SMEs using machine learning technique which would aid in arresting this monster called tax fraud. Simple static equation (SSE) and logistic regression (LR) were used to formulate the model which has a two-layer architecture that uses the sigmoid that takes a real-valued input and outputs a value between 0 and 1. This is research in progress, hence, the researchers are in the process of collecting a historical data for testing the model, meanwhile a questionnaire was sent out prior to the study to ascertain the need analysis of this study.

Keywords: Tax, SMEs, Tax Fraud, Machine Learning Techniques, Detection