Based on the doctor-patient interaction forum of “Haodaifu Online”, “Xunyiwenyao” and “Youwenbida”, taking advantage of the text mining technologies and methods such as Chinese word segmentation, data cleaning and filtering, word frequency statistics, semantic network map and visualization, extract and summarize the symptoms, causes, common drugs, complications, non-drug therapy and nursing method of Deep Vein Thrombosis(DVT), the paper verifies the availability of text mining method in exploring the prevention law and method of diseases and the feasibility of application of doctor-patient interaction Q&A network data source in medical research.