Research On Evaluation Of Special Hospitals Based On The User Reviews On The Internet

With the rapid development of web2.0 and medical informatization, more and more specialized medical service websites arise, such as www.haodf.com, www.guahao.com and so on. Most medical portal websites not only provide professional medical services to users, but also support users to share experience and comment on hospitals and doctors etc. Furthermore, many patients and relatives are willing to share their feelings on the treatment effects and medical staff service attitudes etc. Other patients can also draw lessons from others’ experience to choose a suitable hospital. But users are difficult to digest and assimilate the large amount of reviews. Therefore, this study aims to mine the user-focused hospital features and sentimental tendency using the rich resources of comments and compare the corresponding features of the same type of hospitals so as to provide decision support for the patients.In this paper, we mine the initial features of the hospitals using the association rule and word frequency statistics, and then introduce an authoritative hospital evaluation index system. Next we analyze the similarity between the initial features and index combining with the semantic knowledge to perfect the hospital features and provide patients with a clear and useful feature systems. And then, we will improve the existing emotional dictionary by adding the network emotional vocabulary. Based on the improved emotional dictionary we can judge the emotional polarity of all the features of hospitals so as to support users to choose suitable hospitals.Choosing three tumor specialized hospitals as the experiment object, this study aims to mine the hospital features and analyze sentimental portal. The experiment results are proved to be acceptable and can provide reference for the patients.