Automatic Identification of Equivalence of Concepts in Different Languages for Never-Ending Learning

Abstract

This paper describes the process of automatic identification of concepts in different languages using a base that relies on simple semantic and morphosyntactic characteristics like string similarity, difference in words amount and translation position on dictionary (when exists) and a neural network that has been used as a model of machine learning. All experiments use data that was obtained from a few categories of Read The Web (RTW) project and an endless learning computation system called NELL: Never-Ending Language Learning. The results were compared with dictionary and showed that the introduction of neural network brought a significant gain in the process of equivalence of concepts.