NLP-based support for ontology lifecycle development

This paper describes the implementation of an approach to modelling the dynamics of the propagation of textually derived semantic information, in particular with respect to networked ontologies. On the one hand, new ontologies may be generated automatically from textual data, or existing ontologies may be modified or extended according to new evidence. This can cause problems for large or networked ontologies, where only a small section of the ontology may be modified, and where multiple users may be working with the ontologies simultaneously. Collaborative editing of both documents and ontologies is becoming more widespread, but can bring major problems with change management: both within the ontology itself and between text and ontology. In this paper, we describe a plugin for the NeOn toolkit which uses both automatic and manual methods for generating and modifying ontologies on the fly from unstructured text, and enables two-way ontology lifecycle development.