<![CDATA[The digitalization of Indonesia’s tax administration through the CoreTax Administration System represents a strategic effort by the Directorate General of Taxes to enhance operational efficiency and taxpayer compliance. However, early implementation has faced various technical and administrative challenges, including server instability, longer processing times, and additional overtime for accountants and tax consultants. This study evaluates the effectiveness of CoreTax using a quantitative triangulation approach that integrates questionnaire data with IndoBERT-based sentiment analysis of 10,084 public tweets collected between December 2024 and March 2025. The Wilcoxon signed-rank test indicates a significant decrease in SPT reporting efficiency and an increase in overtime during the transition phase. Sentiment analysis results show a dominant 64.7% negative sentiment that gradually declined over time, reflecting users’ adaptation to the new system. The IndoBERT classification model achieved 91.7% accuracy and a 0.903 macro-F1 score, confirming strong model reliability. Integration of both datasets reveals a consistent interpretation that CoreTax performance improves after initial adjustment. These findings highlight the importance of system quality enhancement and continuous sentiment monitoring to ensure sustainable digital tax adoption in Indonesia.]]>
Copyrights © 2026
