Abstract: To enhance the intelligence level of railway transportation research, this paper designed and implemented an intelligent question and answering system for the field of railway transportation. It elaborated on the system architecture and specific functions, used a locally deployed large language model and integrated natural language processing, structured database retrieval, and semantic vector indexing technologies to implement intelligent querying and path analysis of core business data. In the actual construction process, the paper constructed a semantic vector library based on embedded models, effectively improved the quality of question and answering for unstructured data such as path rules. At the same time, it introduced standardized data processing flow and dynamic update mechanism to ensure the completeness and timeliness of railway transportation volume information. The test results show that the system can significantly reduce the difficulty of researchers in processing complex data, improve user query efficiency, and effectively meet the intelligent question and answering needs in the field of railway transportation.