<![CDATA[The purpose of this study is to analyze the financial transparency of local governments in Indonesia through Internet Financial Reporting (IFR). The approach used is machine learning with the Classification and Regression Trees (CART) algorithm. The data used were secondary data, with 546 local governments in Indonesia as observation units. Data were analyzed using R version 4.5.1. The results of this study indicate 44% or 238 local governments are not transparent in disclosing financial reports through their official websites. The results of the CART analysis show that Audit Opinion is the variable that most influences IFR. Other independent variables also influence IFR, namely Regional Revenue, Financial Independence Ratio, Administrative Age of Local Government, and Per Capita Income. The optimal classification tree produces 7 segments that are able to predict new data with an accuracy of 64.24% and an AUC of 63.83%. The theoretical implications of this research provide a important contribution to the development of public sector accounting theory, particularly within the theoretical framework of transparency, accountability, and good governance. Practical implications include formulating strategies to minimize the number of local governments that are non-transparent in disclosing financial reports on their official websites, the characteristics of which are reflected in segments 5, 6, and 7. The novelty of this research is that it provides new insights into the analytical methods used. The use of machine learning with the CART algorithm has never been used by previous researchers to analyze regional financial transparency through IFR.]]>
Copyrights © 2025
