Now a day people tend to seek knowledge or information from internet that concern with health through online healthcare services. The basic aim of this system is to bridge the vocabulary gap between the health providers by proving instant replies to the questions posted by patients. Automatic generated content for healthcare services are chosen instead of traditional community generated systems because they are reliable, compatible, and provide instant replies. This paper proposes a scheme to code the medical record using local mining and global approaches. Local mining aims to code the medical records by extracting the medical concepts from individual record and then mapping them to terminologies based on  external authenticated vocabularies. Local Mining establishes a tri-stage framework to accomplish this task. Global learning aims to learn missing key concepts and propagates precise terminologies among underlying connected records over a large collection.