Inherent Biases of Recurrent Neural Networks for Phonological Assimilation and Dissimilation Article Swipe

View

Related Concepts

Recurrent neural network Feature (linguistics) Computer science Artificial neural network Dissimilation Consonant Artificial intelligence Feature learning Vowel Representation (politics) Speech recognition Linguistics Politics Political science Philosophy Law

Amanda Doucette ·

YOU? · · 2017 · Open Access · · DOI: https://doi.org/10.48550/arxiv.1702.07324 · OA: W2949879527

A recurrent neural network model of phonological pattern learning is proposed. The model is a relatively simple neural network with one recurrent layer, and displays biases in learning that mimic observed biases in human learning. Single-feature patterns are learned faster than two-feature patterns, and vowel or consonant-only patterns are learned faster than patterns involving vowels and consonants, mimicking the results of laboratory learning experiments. In non-recurrent models, capturing these biases requires the use of alpha features or some other representation of repeated features, but with a recurrent neural network, these elaborations are not necessary.