To explore the patients’ focus under different emotional comments, this paper uses the ROST CM text mining software to build semantic network of high-frequency word co-occurrence to mining patients’ online comments in Good Doctor Online community from two dimensions of positive and negative reviews. Finally, this paper sums up the main factors influencing patients’ satisfaction, including service attitude, technology level, hospital diagnosis and treatment process and patient expectations, and gives suggestions to doctors, hospitals and medical service according to the results of the analysis.