Human-AI Collaboration in Research: A Case Study on RAG-Driven Information Retrieval from Scientific Papers
Disciplines
Artificial Intelligence and Robotics
Abstract (300 words maximum)
Every day AI is becoming increasingly important to more and more industries and applications. AI usage is still relatively new to most fields, and many are unfamiliar with how to achieve the best results, which often require heavy interaction and involvement from users. This human and AI teaming while usually producing the best results is still an underdeveloped point of research which has not been explored enough. This research aims to amend this by exploring papers about interactive AI through the development and usage of a tool to collect and study. The tool is a web app that allows for the search, retrieval, and management of papers to then be explored or “talked to” through an LLM. The LLM portion of the tool will allow for the user to select one or more papers to allow for the user to explore the selected papers in creative ways like comparing and contrasting papers, connecting topics, or just deep diving into a paper. Through our research, we’ve developed an innovative, easy-to-use prototype web app that allows users to dynamically engage with scientific papers. This tool has facilitated further research on human-AI collaboration and serves as an example of how humans and AI can collaborate.
Academic department under which the project should be listed
CCSE - Computer Science
Primary Investigator (PI) Name
Md. Abdullah Al Hafiz Khan
Human-AI Collaboration in Research: A Case Study on RAG-Driven Information Retrieval from Scientific Papers
Every day AI is becoming increasingly important to more and more industries and applications. AI usage is still relatively new to most fields, and many are unfamiliar with how to achieve the best results, which often require heavy interaction and involvement from users. This human and AI teaming while usually producing the best results is still an underdeveloped point of research which has not been explored enough. This research aims to amend this by exploring papers about interactive AI through the development and usage of a tool to collect and study. The tool is a web app that allows for the search, retrieval, and management of papers to then be explored or “talked to” through an LLM. The LLM portion of the tool will allow for the user to select one or more papers to allow for the user to explore the selected papers in creative ways like comparing and contrasting papers, connecting topics, or just deep diving into a paper. Through our research, we’ve developed an innovative, easy-to-use prototype web app that allows users to dynamically engage with scientific papers. This tool has facilitated further research on human-AI collaboration and serves as an example of how humans and AI can collaborate.