<![CDATA[Public interest in financial technology, especially in digital loans, has grown rapidly in Indonesia over the past five years. This study analyzes weekly Google Trends data from June 2020 to June 2025, focusing on the keywords “pinjaman online” and “pinjol.” The research aims to understand usage patterns, identify interest spikes, and build predictive models for public attention using ARIMA and various machine learning methods. The ARIMA model, used as a baseline, produced a Mean Absolute Error (MAE) of 2.88 and Root Mean Square Error (RMSE) of 3.61. In contrast, the XGBoost model yielded superior accuracy, with an MAE of 1.39 and RMSE of 2.03. Among all models tested, XGBoost most accurately captured the volatility of online interest, especially during peak periods like October 2021 and January 2023. Findings confirm that “pinjol” outperforms “pinjaman online” in search frequency, driven by informal usage and media coverage. These insights have implications for financial regulators, fintech marketers, and social scientists studying digital behavior, suggesting machine learning models provide a more reliable approach to forecasting online public interest.]]>
Copyrights © 2025
