This paper briefly summarizes the characteristics of a computer-based medical terminology constructed by the author. Its size is around 140,000, systematized by meaning using a single attribute as key: similarity among instances covered by the term. This systematization resulted in a single hierarchical tree. Initial term typing was based on the knowledge patterns of the terms, represented by the arcs (relationships) among terms (nodes). Six different types of medical terms could be identified. For coding of a medical term, its position on the hierarchical tree was used, the semantic address. It is recommended that the semantic address should be used for worldwide compatibility.