<![CDATA[In today's digital world, information technology has had a significant impact on various aspects of life, including the study and exploration of Hadiths. This research aims to develop a Hadith question answering system that enables users to obtain quick and accurate answers to Hadith-related questions. By utilizing LangChain and Gemini-pro models, and implementing the merger retriever method, this system provides an innovative solution for searching Hadiths from the nine main books (kutubut tis’ah). LangChain and Gemini-pro ensure that the generated answers are not only relevant but also accurate, utilizing Large Language Model (LLM) technology. Additionally, the implementation of the merger retriever enhances the system's performance in finding the most suitable answers from various Hadith sources simultaneously. This research involves a series of stages ranging from collecting Hadith data, system analysis and design, implementation, to system testing. The test results show that the Hadith question answering system can produce relevant and accurate answers, with expert Hadith respondents evaluating the quality of answers averaging 88.5% satisfaction with the "Strongly Agree" rating interval, and LangChain evaluator scoring averaging 91% within the "Strongly Agree" category interval. This emphasizes the significant potential of applying information technology in the study of Hadiths, and opens opportunities for further development in this field. This research is expected to make a significant contribution to the ease of access to accurate and fast Hadith information for the Muslim community.]]>
Copyrights © 2024
