Digital Library

cab1

 
Title:      A WEB-BASED METHOD FOR ONTOLOGY POPULATION
Author(s):      Hilário Oliveira, Rinaldo Lima, João Emanoel, Fred Freitas
ISBN:      978-989-8533-09-8
Editors:      Bebo White and Pedro Isaías
Year:      2012
Edition:      Single
Keywords:      Ontology Population, Confidence Metric, Pattern Learning, Similarity Measure.
Type:      Full Paper
First Page:      85
Last Page:      92
Language:      English
Cover:      cover          
Full Contents:      click to dowload Download
Paper Abstract:      The manual development or improvement of ontologies is a very time-consuming and error-prone task. Thus, an automated or semi-automated mechanism able to extract instances of both concept and relations is highly desired. In this paper, we present our ongoing research work on Ontology Population (OP) from the Web. We describe an unsupervised method for OP that takes profit of the high redundancy present in the Web. The system prototype also described in this paper implements our autonomous domain-independent method for extracting instances of concepts defined by a domain ontology. We focus on the proposal and evaluation of a new version of a confidence-weighted metric proposed in previous work. In addition, we analyze the application of a Part-of-Speech filtering mechanism that seems to improve the earlier version of our algorithm for learning effective surface text patterns. The achieved results so far are very encouraging.
   

Social Media Links

Search

Login