Digital Tax Filing Compliance Analysis Using Pattern Recognition

Dhanaraj Sathiri

Abstract: Research focusing on tax compliance investigations is often anecdotal. In this study, data from the Thai Revenue Department's electronic tax filing system are used with pattern recognition techniques to examine the compliance behavior of both businesses and individual income tax filers. The empirical investigation seeks to identify patterns of behavior and any other anomalies. Importantly, the investigation is not undertaken to explain causes and consequences of tax compliance but simply to provide a framework for investigation and a database based around a pattern recognition analysis. The analysis reveals clear patterns of compliance behavior within the two data sets, although the data from businesses is more consistent. Anomalies and irregularities are also evident, often recommending further and deeper investigation. Such an approach to tax compliance investigations provides broad and objective insight into typical behavior, while highlighting specific sets of records that require further investigation. Although these insights into the compliance behavior of Thai tax filers are somewhat limited, they nonetheless have potential value for the country's tax administrators. The framework can, however, be deployed in more general research into electronic tax filing compliance and enables the recognition of behavior that is not only anomalous but also worthy of further scrutiny. This enables, and possibly improves, resource allocation by tax authorities seeking to maximize the benefits from compliance investigations while minimizing the costs.

Keywords: Tax Compliance Investigations, Electronic Tax Filing Systems, Pattern Recognition Techniques, Compliance Behavior Analysis, Business Tax Filers, Individual Income Tax Filers, Anomaly Detection in Tax Data, Empirical Tax Analytics, Revenue Department Data, Behavioral Pattern Identification

How to Cite?: Dhanaraj Sathiri, "Digital Tax Filing Compliance Analysis Using Pattern Recognition", Volume 11 Issue 12, December 2022, International Journal of Science and Research (IJSR), Pages: 1524-1531, https://www.ijsr.net/getabstract.php?paperid=MS2212181744, DOI: https://dx.doi.org/10.21275/MS2212181744