Approach to automatic population of ontologies of scientific subject domain using lexico-syntactic patterns

Abstract

Abstract This paper presents an approach to automatic population of ontologies of a scientific subject domain (SSD) using Lexico-Syntactic Patterns (LSPs) and a corpus of texts related to modeled domain. The main feature of this approach is that such patterns are automatically built based on Ontology Design Patterns of other types provided by the system for the automated development of SSD ontologies using heterogeneous Ontology Design Patterns. The implementation of the ontology population using constructed LSPs is described in detail. The results of the experiments on the SSD ontology population are presented. It is noted that there is a problem in establishing a subject of a relation when extracting facts. To address this problem, the authors are planning to employ the coreference resolution methods.