NAMED ENTITY DISAMBIGUATION BASED ON ONTOLOGY DATA INTERPRETATION

Antonova Elena Sergeyevna

Information About Author:

The article deals with the issue of named entity disambiguation (geographical objects in English news). The research is based on the method of ontology data interpretation. The method uses features, extracted from Freebase ontology, as the input for a neural network (multi-layer perceptron). This strategy makes possible to achieve precision value of 95.5 % and recall 92.9 %.

Keywords: named entity recognition, named entity disambiguation, neural networks, ontology

( 475.54 kB ) ( 460.7 kB )

Issue: 8, 2013

Series of issue: Issue 8

Rubric: INTERDISCIPLINARY STUDIES

Pages: 200 — 204

Downloads: 918