Multi Agent Based Intelligent Personal Research Assistant

Since the very early days, developments of mankind have been energized by research. Despite people having developed many research methods, including Scientific Method, Milestone Approach, and IMRD, research conduct remains a challenge. We have examined various research methods and discovered reading and writing as the foundation of all research methods regardless of a novice or an experienced researcher. This was practised even before formally acknowledging what we called research methods today. We developed an Intelligent Personal Research Assistant, InPRA, which guides a researcher in systematically reading past research papers and, more importantly, developing research proposals, progress reports, thesis, and papers while doing the research, without waiting for the last moment. While InPRA is based on the reading-writing model of research, within InPRA, a researcher can access research management tools, scientific writing tools, specific research methods and other resources. InPRA has been developed as Multi Agent System to enable interaction among the steps in the research process beyond sequential processing. Research students and experienced researchers like supervisors have evaluated the guidance of InPRA to generate material for a project proposal, which is the first document in a research process. According to the evaluation results, the multi-agent-based approach in InPRA could outperform traditional text-based search approaches. Moreover, based on survey results, the system is very useful for researchers as it generates more relevant results.