<![CDATA[The rapid development of social media creates enormous volumes of text data, requiring sentiment analysis tools that are easily accessible to the public. The purpose of this community service is to develop the DataCraft application as a free web platform for social media sentiment analysis using machine learning algorithms. The implementation method includes a 1-month data collection phase and 1.5 months application development using Python Flask, Naive Bayes, and Support Vector Machine (SVM) technologies. The application is equipped with data exploration, automatic preprocessing, modeling, and result visualization features. Test results show Naive Bayes accuracy reaches 77.25% with a user-friendly interface. The application successfully provides free access to students, researchers, and practitioners to perform sentiment analysis without requiring high programming skills. The impact of community service shows an increase in digital literacy capabilities of the community in analyzing public opinion on social media. DataCraft is available open-source on GitHub with complete tutorials to facilitate wider adoption.]]>
Copyrights © 2025
