Development of a Library Information Chatbot for Lampung University Based on Natural Language Processing

Puput Budi Wintoro (Universitas Lampung)
Khairudin (Universitas Lampung)
Zulmiftah Huda (Engineering Faculty, Universitas Lampung, Bandar Lampung, Lampung, 35141, Indonesia)
Rio Ariestia Pradipta (Engineering Faculty, Universitas Lampung, Bandar Lampung, Lampung, 35141, Indonesia.)

Proceedings of International Conference on Multidisciplinary Engineering (ICOMDEN)
Vol. 2 (2024): Proceedings of International Conference on Multidisciplinary Engineering (ICOMDEN)

The University of Lampung (Unila) Library has over 86,783 registered patrons, with 12,329 active users, including faculty, students, staff, and external patrons. Services include verification, circulation, procurement, e-books, and journals. However, limited staff availability to respond to inquiries has negatively impacted service satisfaction. To address this issue, the implementation of chatbot technology is proposed as a solution. A chatbot simulates human conversation through text, voice, or visuals. There are two types: Flow-Based Chatbots, which follow a predetermined conversation flow, and Open-Ended Chatbots, capable of handling dynamic conversations. Development methods include Fixed Rule-Based and Machine Learning/Natural Language Processing (ML/NLP) Based Chatbots. This research aims to develop a Flow-Based Chatbot using ML/NLP on the Dialogflow platform, integrated with Unila Library's local database through a Python-based API, specifically FastAPI. The implementation of this chatbot is expected to enhance the responsiveness and availability of library services, ultimately increasing patron satisfaction.