The invention provides a
natural language-driven scientific research naming and
ranking patient matching method and
system, belongs to the technical field of
artificial intelligence and
medical information processing, and solves the problems of high manual
workload, difficulty in semantic analysis, insufficient
logical reasoning ability and the like in an existing
cancer patient screening mode. The method comprises the following steps: analyzing a containing and arranging condition in a
natural language text form in the tumor field to obtain a structured containing and arranging condition logic rule set; performing entity recognition, relation extraction and portrait construction on the multi-
modal medical record of the
cancer patient to obtain a comprehensive
cancer portrait of the patient; and performing logic matching on the comprehensive cancer portrait of the patient and the structured accommodating and discharging condition rule set, determining whether the patient meets accommodating and discharging conditions, and generating a traceable matching interpretation path. The method can effectively solve the problems of large manual
workload, difficulty in semantic analysis, insufficient
logical reasoning ability and the like in the existing cancer
patient screening mode.