BUILDING A NEWS-SHARING WEBSITE COMBINED WITH LLM FOR INFORMATION RETRIEVAL

Abstract

In recent years, there has been significant development in machine learning and large language models, such as ChatGPT. The growing demand for utilizing these models to answer questions based on document or website data is evident. However, it is important to note that large language models are trained with a substantial dataset at a specific point in time, limiting their responses to information available before that training period. This poses a challenge when attempting to employ these models for retrieving recent information. Furthermore, in enterprise environments, there is a reluctance to expose organizational information outside the company, leading to restrictions on employees using cloud services like ChatGPT. Additionally, students in software engineering majors may find it challenging to integrate ChatGPT into software applications due to its novelty and cost, creating a potential barrier for some learners. Recognizing these challenges, the primary objective of this thesis is not only to develop an application for sharing information but also to address user queries based on real-time data without incurring fees and with the capability to run on local computers.