Healthcare professionals need to keep themselves updated with the latest medical developments by finding and reading relevant articles in order to provide the best possible care to their patients. The most popular technique for retrieving relevant articles from a digital library is keyword matching, which is known to retrieve a large amount of irrelevant articles without taking into account the knowledge requirements the user. Currently, the research community is making progress, but is still far from resolving this problem. In this paper, we propose a new method for generating rule-based stereotypical profiles to capture the knowledge requirements based on user roles, and an information filtering technique that combine content-based and rule-based filtering to deliver relevant articles to a user.