Comparing and consolidating two heuristic metaschemas.

A metaschema is an Abstraction Network of the UMLS's Semantic Network (SN) obtained from a connected partition into its collections of semantic types. A metaschema can support comprehension, visualization and navigation of the SN. In this paper, a methodology is presented to evaluate and compare two given metaschemas, based on their structural properties. This methodology was applied to the top-down metaschema and the bottom-up metaschema, derived from two studies involving two groups of UMLS experts. By the Cronbach alpha estimate, the bottom-up metaschema is reliable, while the top-down metaschema is not. An algorithm was designed to yield a consolidated metaschema enjoying the best and avoiding the worst of the two given metaschemas. The consolidation algorithm was applied to the above two metaschemas, The resulting consolidated metaschema has better structural properties than either of the two metaschemas and better supports user orientation visualization, and navigation of the Semantic Network.