CHATBOT PERENCANAAN PEMBANGUNAN KABUPATEN JAWA BARAT DENGAN SUPPORT VECTOR MACHINE DAN TF-IDF

Abstract

Rapid population growth and development inequality in West Java Province are challenges for local governments. Therefore, this research develops a development planning Chatbot as a tool to reduce the workload in handling repetitive requests in the regional development planning process in Kuningan Regency as a sample category of areas with low population density. The system is designed using Term Frequency-Inverse Document Frequency (TF-IDF) method to process user messages into numerical matrix, Support Vector Machine (SVM) method to classify user questions automatically, and Streamlit framework for the interface. The developed chatbot is able to generate reports on economic growth trends based on the value of gross regional domestic product at constant prices and business sector recommendations for investors. Chatbot test results in answering valid or correct questions amounted to 60%. Although the accuracy still needs to be improved, this system can still be a useful tool in providing important information for local government decision-making in development planning in low population density districts in West Java.