AI-Based Chatbot System for Education and Recommendations on the Use of Native Papuan Herbal Plants using the Large Language Models method

Suhardi Aras (Program Studi Teknik Informatika, Universitas Muhammadiyah Sorong, Indonesia)
Maya Nurliati Kelian (Program Studi Teknik Informatika, Universitas Muhammadiyah Sorong, Indonesia)
Ilham S (Program Studi Teknik Informatika, Universitas Muhammadiyah Sorong, Indonesia)
Alfiyyah Faridah (Program Studi Teknik Informatika, Universitas Muhammadiyah Sorong, Indonesia)

International Journal of Informatics, Information System and Computer Engineering (INJIISCOM)
Vol. 6 No. 1 (2025): INJIISCOM: VOLUME 6, ISSUE 1, JUNE 2025

This study develops an AI-based chatbot to educate users and recommend native Papuan herbal plants, aiming to preserve traditional knowledge that is poorly documented in local communities. The methodology includes problem analysis, literature review, data collection, and system implementation using Google Generative AI (Gemini-Pro) as a Large Language Model (LLM). The system analyzes text queries to provide relevant answers. Evaluation using the BERTScore method showed good performance with an average Precision of 0.58, Recall of 0.67, and F1-Score of 0.62. The higher recall demonstrates the system's effectiveness in retrieving pertinent information. This research highlights the potential of AI to enhance public understanding of Papuan local wisdom and support traditional herbal medicine.