The Design And Implementation Of Context-Aware Recommendation System On Hospitalized Information

With the growing use of smart phones, more and more hospitalize information platform will gradually apply its partial information service into mobile applications. Meanwhile, users are faced with the increasingly serious problem of “overload information” because of the growing number of medical information, then recommendation system came into being. In the medical treatment information platform, the traditional recommendation algorithms could not completely solve cold starting problem for new users. In addition, recommendation system based on location could not achieve the goal of personalize recommendation.Based on the above issues, this paper proposed a hospitalize information recommendation system which merged a variety of context information by deeply studying in context-aware and recommendation algorithms. The application will use baidu map to obtain users’ context information, exploit ontology to model context information, combine context information and properties to get user’s interest, finally employ recommendation algorithms based on rules to recommend hospitals and doctors for the user. Thus, patients will obtain personalize recommendation in seeking medical advice by this application.In this paper, the main work is as follows:1. Analyze various types of contextual information for hospitalize environment, such as time context, location context, user context and other context. Then use ontology to model for these context information.2. Design for this recommendation system, including the need for system, system architecture, system function module and database.3. Analyze and Design recommendation subsystem, including model for user interests, select algorithm, design system framework, prompt its processes, design rule base and calculate by these rules.4. Realize all the functional modules, including context-aware modules such as context information obtaining and ontology inferring, core modules such as two kinds of recommendation, public modules such as login and registration, other modules such as questions asking, electronic medical records adding and hospital details.The innovation of this paper are the following three points:1. According to the characteristics of mobile applications, this paper will use recommendation algorithms based on rules to effectively solve cold start problems faced by the new users and enable new users to spend less learning cost to obtaining recommended information.2. Merging a variety of special contexts into recommendation and extending the recommendation based on location to context awareness will provide users with more personalized recommendations.3. Applying the recommendation system mixed a variety of context information into the mobile platform for hospitalize information will promote the development of hospitalize information system in the mobile applications, be easier for users to practice medicine and hospitalize, and help to solve the multiple problems faced by the current wisdom medical.