Analysis of Agenda Setting and Sentiment in Regional Revenue and Expenditure Budget Reporting on Kompas.Com for the 2020–2024 Period Using Computational Content Analysis

This study aims to analyze the pattern of reporting on the Regional Revenue and Expenditure Budget (APBD) in the online media Kompas.com during the 2020–2024 period using a computational content analysis approach. This study focuses on the distribution of intensity, sentiment tendencies, and themes and narrative constructions built by the media related to the APBD issue. Data were taken through web scraping of 590 articles containing the keyword "APBD". The analysis was carried out in stages, starting from data cleaning, descriptive analysis of article distribution and word frequency, temporal analysis, to sentiment analysis using an Indonesian language machine learning model and LDA-based topic modeling. The results of the study show that the intensity of APBD reporting is fluctuating and is greatly influenced by political moments and crises, such as the COVID-19 pandemic. The dominant sentiment is neutral (94.9%), with the dominant themes of evaluation, supervision, and transparency in the reporting. The main topics found include budget realization and absorption, political dynamics of APBD discussions, socio-economic impacts, institutional supervision, and policy responses to the pandemic. This finding confirms that online media such as Kompas.com not only play a role as a conveyor of information, but also build framing and agenda setting in regional fiscal policy issues, which have implications for the formation of public perceptions of regional financial governance in Indonesia.