目的 研究中医智能问诊系统，实现快速获取关键症状并完成辨证，为中医问诊智能化、客观化提供了一种新的思路和方法。方法 采用基于物品的协同过滤推荐算法（Item-Based Collaborative Filtering，ItemCF）和遗传算法构建症状获取模块以获取患者的症状，利用随机森林算法构建分类器并基于获取到的症状完成中医辨证。结果 该系统实现了高效地获取患者症状并完成中医辨证。在13次提问次数下，便能获得辨证所需的核心症状，实现证候分类器90%以上的辨证效果。结论 该问诊系统能够较好地解决中医问诊中“问什么、怎么问”的两个核心问题，相比依据问诊量表获取症状，极大地简化了问诊中关键症状获取的过程，并能够在证候分类中保持较好的分类效果，在问诊客观化研究上具有一定的实用价值。

Objective To study the intelligent diagnosis system of traditional Chinese medicine (TCM), realize the rapid acquisition of key symptoms and complete syndrome differentiation, and provide a new way of thinking and method for the intellectualization and objectification of TCM diagnosis.Methods The item-based collaborative filtering (ItemCF) and genetic algorithm were used to construct the symptom acquisition module to obtain the symptoms of patients. The random forest algorithm was used to construct the classifier and complete the syndrome differentiation based on the acquired symptoms.Results The system can effectively obtain the symptoms of patients and complete TCM syndrome differentiation. The system can obtain the core symptoms needed for syndrome differentiation and achieve the syndrome differentiation effect of more than 90% of the syndrome classifiers under 13 inquiries.Conclusion The experimental results show that the inquiry system can better solve the two core problems of “what to ask and how to ask” in the TCM inquiry. Compared with obtaining symptoms according to the inquiry scale, it greatly simplifies the process of symptom acquisition in the inquiry and can maintain a better classification effect in the syndrome classification. This research provides a new idea and method for the intellectualization of TCM inquiry, and has a certain practical value in the objectification research of inquiry, and has a certain practical value in the objectification research of inquiry.

R241.2