Year: 2006
Type: article
Abstract: Several algorithms for learning syntactic categories from distributional information were tested against utterances from adults and children in twelve typologically different languages.The evaluation measure that was developed allows one to examine word order constraints over a whole corpus and developmentally.By comparing several different algorithms of varying abstraction against actual corpora of children's speech, the evaluation measure determined that lexically specific knowledge is more advantageous than more broad-based category knowledge in predicting word order.
Institution: Max Planck Society
Language: English
Subfield: Developmental and Educational Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W2624616057
