Contextual hybrid-based recommendation of PUBMED articles

ABSTRACT

Background:The amount of information at the medical field has been growing day by day. Also new medical articles about the preoccupied disease are published each day and the updated information is very required for physicians. Consequently,the appropriate information in the appropriate moment is very suitable for physicians, and this is the challenge contributed in this research work. Methods:Our goal is to recommend documents deemed relevant to doctors, moreover, regarding the context of using of an Electronic Medical RecordsEMR. The principle is to extract the context of this usage illness, Age ..., searching on the contents of documents and taking into account the rate of vote documents. For experiment and evaluation, we have used 100 articles randomly selected from PUBMEDabout cardiology. In addition, we have developed a system that extracts the context of EMRat the moment of exploration. The extracted context is used with users rating bythe recommender system to select and rank the recommended articles for physicians in the same moment of use.Results:The first result of this research work is the smart interaction between users and the software system by introducing the context of use. In addition, another important result is the reuse of user's appreciation for additional dynamicity and intelligibility. Conclusion: The developed system offers the physician an appropriate recommendation of pertinent PUBMED articles. The developed system augments the relevancy of the recommendation by analyzing the contents of the articles and introducing a collaborative method: