Abstract
The automatic ontology enrichment consists of automatic knowledge extraction from texts related to a domain of discourse in the aim to enrich automatically an initial ontology of the same domain. However, the passage, from a plain text to an enriched ontology requires a number of steps. In this paper, we present a three steps ontology enrichment approach. In the first step, we apply natural language processing techniques to obtain tagged sentences. The second step allows us to reduce each extracted sentence to an SVO (Subject, Verb, and Object) sentence, supposed to preserve main information carried by the original sentence(s) from which it is extracted. Finally, in the third step, we proceed to enrich an initial ontology built manually by adding extracted terms in the generated SVO as new concepts or instances of concepts and new relations. To validate our approach, we have used “Phytotherapy" domain because of the availability of related texts on the WWW and also because its usefulness for pharmaceutical industry. The first results obtained, after experiments on a set of different texts, testify the performance of the proposed approach.
Highlights
Ontology allows knowledge representation in graphical and intuitive manner but its construction and management is a hard task and a very time consuming operation
This maintaining operation is sometimes called enrichment, sometimes it is called population as well as, but what is exactly the precise meaning of each one of this words? Ontology population is the process of inserting concept and relation instances into an existing ontology while ontology enrichment is the process of extending ontology, through the addition of new concepts, relations and rules [15]
In the context of this paper, to the ontology enrichment process covering the three first steps of the ontology learning process, where we propose an approach for automatic ontology enrichment giving a text relating to a target domain
Summary
Ontology allows knowledge representation in graphical and intuitive manner but its construction and management is a hard task and a very time consuming operation. With the apparition of internet and new information and communication technologies, the mass of produced texts relating to different domains becomes huge and almost available for exploitation by interested users. It would be very useful if this maintaining operation of ontologies will be done in an automatic or semi-automatic manner. Ontology population is the process of inserting concept and relation instances into an existing ontology while ontology enrichment is the process of extending ontology, through the addition of new concepts, relations and rules [15]. The ontology learning process is composed of several steps which are concept learning, taxonomic relation learning, non-taxonomic relation learning and axiom and rule learning
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
