Word Sense Discrimination Using Statistic Analysis of Texts

Authors

  • Rogelio Nazar Universitat Pompeu Fabra

https://doi.org/10.4471/brac.2013.01

Keywords:


Downloads

Abstract

For years, computer programs have been working to obtain information about certain entities such as persons, organizations or scientific concepts from the Web or from other sources. However, they have many challenges yet to overcome, for instance when texts refer to different entities that share the same name (e.g., a mouse can be an electronic device or a living creature). This article presents a method to solve this problem based on the frequency analysis of the words that are found in the vicinity of a target word. Each sense of the polysemous word or term will be represented as a different group of other vocabulary units that show a tendency to appear together with the target word in each of its different senses. The interest of the proposal is that it does not require previous knowledge about the language of the corpus or any other formof knowledge from the external world.

Downloads

Download data is not yet available.

Metrics

Metrics Loading ...

Downloads

Published

2013-06-03

Almetric

Dimensions

How to Cite

Nazar, R. (2013). Word Sense Discrimination Using Statistic Analysis of Texts. Barcelona Research Art Creation, 1(1), 5–26. https://doi.org/10.4471/brac.2013.01

Issue

Section

Articles