Ontology of Properties and its Methods of Use: Properties and Unit extraction from texts

摘要

Methods of automatic extraction and identification of properties, quantities and units of measurement from texts are considered. The developed ontology of properties and units of measurement includes both fundamental connections of properties and units of measurement and connections that show the formation of concepts. The technology of property identification is based on fact that semantically significant text elements extracted within the boundaries of one or several sentences forming the semantic neighborhood of the property are correlated with the corresponding components of the ontology, which allows to restore the missing semantic fragments or identify discrepancies in the designations. The results of experimental studies of the effectiveness of the developed tools are presented.