Output details

A Probabilistic Model of Syntactic and Semantic Acquisition from Child-Directed Utterances and their Meanings

<22>Originality: First broad-coverage model of syntactic/semantic acquisition (evaluated on real child-directed language); previous models were evaluated on theory-internal criteria or used artificial corpora.

Significance: Some researchers argue that statistical learners cannot exhibit the effects mentioned under Rigour (Thornton and Tesan 2007); this work shows them wrong, providing important support for statistical models of child learning. It also introduces a new data set which will be useful for other researchers to test their models.

Rigour: The CCG framework and online probabilistic learning algorithm allow the model to capture important effects in child language acquisition, including 'fast-mapping' and sudden jumps in performance.