| Title |
Multilingual OTC Drugs Chatbot System Using Retrieval-Augmented Generation |
| Authors |
조항준(Hangjun Jo) ; 김서영(Seoyeong Kim) ; 정대원(Daewon Jeong) ; 오희빈(Heebin Oh) ; 정경용(Kyungyong Chungark) |
| DOI |
https://doi.org/10.5370/KIEE.2025.74.11.1972 |
| Keywords |
Fine-tuning; Prompt-tuning; Retrieval-Augmented Generation; Chatbot; OTC Drugs Information Provision Chatbot |
| Abstract |
In this paper, we propose the multilingual Chatbot-based OTC drugs information summary model using retrieval-augumented generation. In this paper, we propose the multilingual Chatbot-based OTC drugs information summary model using retrieval-augumented generation. In this paper, we propose the multilingual Chatbot-based OTC drugs information summary model using retrieval-augumented generation. In this paper, we propose the multilingual Chatbot-based OTC drugs information summary model using retrieval-augumented generation. The proposed method use to provide accurate answers to OTC Drugs questions to users using OpenAI's GPT 3.5 model.
The GPT 3.5 model, which was tuned prompts to provide answers based on official data, was used. In addition, the model was trained to answer with a consistent tone through fine tuning. And the research was conducted to provide real-time document-based answers through RAG-type research. This paper explains the design process of the chatbot system based on OpenAI's GPT 3.5 model and proposes a method of providing over-the-counter information more accurately and effectively in various languages. This study is expected to increase the reliability and accessibility of providing over-the-counter information to multinational users staying in Korea. In addition, it is expected to be able to provide answers based on various real-time information. |