Human-AI Collaboration Empowered Knowledge-Oriented Agent in Large Language Model

This study proposes a Human-AI collaboration knowledge-oriented agent based on Large Language Model to address the core problems of low level of standardization, high subjectivity and inefficient knowledge transfer in the traditional bid evaluation in the construction industry. Through Design Science Research Methodology, Knowledge Graph and actual enterprise case, an intelligent agent for bid evaluation is constructed. Through automatically parsing massive bidding text data, the agent realize the tasks of structured extraction, comparison and scoring of evaluation items, and forming a dynamic feedback by combining the empirical knowledge of business personnel. In addition, the study proposes "Human-AI Collaboration" decision-making which not only preserves the core value of human experience, but also realizes the iterative optimization of knowledge exploration. This study provides a theoretical framework and practical path for the intelligent transformation of the construction industry, and also provides a reference for the design of Human-AI Collaboration mechanisms in LLM applications.