<![CDATA[Indonesia's economic transformation requires industrialization supported by the capital market as a catalyst. The capital market plays an important role in the economy by relying on information from various sources, including online news, which affects the movement of the JCI. This study aims to detect JCI movement based on online news using transformer models, namely indoBERT-base, indoBERT-large, indoBERT-lite-large, and multilingual-BERT. Data was collected through web scraping from the detik.com page from August 2018 to May 2024. Machine learning models such as Random Forest, Support Vector Machine, and Naive Bayes are used as baseline models. The results show that indoBERT-large and multilingual-BERT have the best performance with an accuracy of 88,60 percent and 95,90 percent. Filtering irrelevant news significantly improves model accuracy. The study concludes that transformer models effectively detect JCI movements, enabling investors to make faster and more accurate decisions, thereby supporting better stock market performance and sustainable economic development.]]>
Copyrights © 2024
