Semi-automatic Approach for Ontology Enrichment Using UMLS

Abstract Ontology enrichment is a process of embedding metadata associated with concepts described in the ontology. Manual information retrieval and enrichment process is labor-intensive and time consuming as each concept is unique and has domain specific meanings. An approach to deal with this problem is to have a unified resource and an automated solution. Different approaches have been used to automate the enrichment process with varying success. Here, we describe our approach of combining automated information retrieval with manual enrichment of retrieved results. Unified Medical Language System implemented on MySQL server was used as a resource for ontology enrichment. To automate the task of information retrieval, KNIME was used which is a workflow management program. The deployed system allows quick retrieval of metadata associated with nearly 1000 ontology terms in a reasonable time frame. Performance evaluation indicated that most of the retrieved results were accurate.