An Ontology Maintenance Approach Based on Collaborative Information Sources and Clustering Techniques
Flávio Ceci, Fernando Benedet Ghisi, Denílson Sell, Alexandre Leopoldo Gonçalves · 2011
Considering that much of the knowledge available on organizations or on the Web are represented on text documents, this work describes a novel approach to support the information extraction and entity recognition from texts in order to leverage ontology maintenance. Our approach applies clustering techniques to speed up the entity recognition phase. For the phases of validation and classification of instances found, we propose the use of collaborative knowledge bases, as well as the help of domain experts, in a semi-automatic fashion. In order to demonstrate the feasibility of the proposed model, an ontology maintenance scenario based on textual descriptions of the courses of an university is presented.