Research On Context-aware Personalized Approach For Recommendation Of E-health Services

The widespread problem of the explosion of service information makes people feel extremely confused and perplexed to choose appropriate services,due to the development of Internet services,which also offers much convenience to people.With the ubiquity of mobile devices such as smart phones,people are able to access Web services at anytime and anywhere.At the same time,the continuous development of communication technology and sense technology makes capturing the contextual information of users easier,which also makes context-aware service attract a large number of researchers at home and abroad in recent years.However,a matter of fact the mobile users can not neglect is the limit to the battery power of mobile devices.People prefer useful and energy-saving services.So,this paper,focusing on the field of e-health service,proposes a new context-aware approach for recommending personalized e-health service,conducting clustering analysis with Rough Set Theory along with analysis of user similarity,location similarity and affiliation based on general contextual information and location.The research work and main contribution of this paper are concluded as follows:1.Aiming at the diversity of user's preference and the data sparsity in the recommendation process of collaborative filtering,an improved user clustering algorithm based on rough set is proposed.Rough set theory is used for rough clustering analysis to find the initial Nearest Neighbor Set of the target user.Then the further recommendation is carried on this basis.The result of the experiment proves that this method is better than the traditional one without clustering.2.Aiming at the general contextual information of the user,a context-aware recommendation model of e-health service in general environment is proposed.In recent years,most e-health-related researches have taken the relationship between doctors and patients into account,while ignoring the contextual information,such as the user's time,mood,etc.However,the contextual information may have an impact on the users' health service preferences.So a context-aware recommendation model of e-health service in general environment is proposed,which incorporates the user's contextual information into the recommendation system,in order to provide more personalized recommendation for the user.3.A location-based approach and an energy-saving approach are proposed forrecommendation of e-health services for users,based on the context-aware recommendation in the general environment,which takes account of the user's location and the energy consumption of the mobile device during the service invocation process in a mobile environment.4.Aiming at the general contextual information and the mobile contextual information of the user,a context-aware approach is proposed for recommendation of e-health services for users and a personalized prototype recommendation system for e-health service is constructed with a case study to validate the proposed approach.In this paper,the prototype system developed based on the relevant theoretical researches can be applied in the actual personalized e-health service recommendation environment.On the one hand,our system takes account of the influence of the e-health service user's gender,age,time and location compared with the traditional system,so it can recommend the personalized service for the user more precisely.On the other hand,our system also considers the energy-saving problem in the recommendation of e-health service,which calculates the energy consumption during the process of the service invocation of the user.So it can recommend personalized and energy-saving health service for the user.