Research on the Intelligent Analysis Method of Main Network Project Cost Based on Large Models and Domain Knowledge

With the increasing complexity of engineering projects, traditional cost analysis methods are no longer sufficient to meet the growing demands. This study proposes an intelligent cost analysis system based on the integration of large language models and engineering cost knowledge graphs, aimed at improving the accuracy and interpretability of main network project cost predictions. By incorporating a knowledge graph-enhanced semantic reasoning mechanism and leveraging the text parsing capability of large language models, the system can extract key information from unstructured engineering documents and perform cost prediction through structured data. The research designs a text parsing process based on language models to achieve efficient conversion of engineering documents into structured data, while using graph reasoning mechanisms to enhance the model's handling of domain-specific knowledge. This addresses the issues of accuracy and interpretability in traditional methods. Experimental results show that the intelligent analysis system performs excellently in cost prediction accuracy, with errors controlled within 7%, and can effectively handle challenges related to missing information and projects of varying scales. This research provides a new direction for the development of intelligent cost analysis methods, with significant practical application value.