A Cognitive-Neuro Computational Lexical Acquisition Model

Abstract

This paper proposes an automatic lexical acquisition model which reflects the characteristics of human language acquisition. The proposed system automatically builds two kinds of lexicon, full-form lexicon and morpheme lexicon by using large corpus as its input. The model is independent of language from which it acquires the lexical knowledge. As experimental results using Korean Sejeong corpus of which size is 10 million Eojeols, the proposed system acquired 2,097 full-form Eojeols and 3488 morphemes. The accumulated frequency of the acquired full-form Eojeols covers the 38.63% of the input corpus and accuracy of morpheme acquisition is 99.87%.