Fractal unfolding: a metamorphic approach to learning to parse recursive structure

Authors:
Whitney Tabor;Pyeong Whan Cho;Emily Szkudlarek
Affiliations:
University of Connecticut, Storrs, CT;University of Connecticut, Storrs, CT;University of Connecticut, Storrs, CT
Venue:
CMCL '12 Proceedings of the 3rd Workshop on Cognitive Modeling and Computational Linguistics
Year:
2012

Citing 9
Cited 0

Learning internal representations by error propagation

Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
Distributed Representations, Simple Recurrent Networks, And Grammatical Structure

Machine Learning - Connectionist approaches to language learning
Analog computation via neural networks

Theoretical Computer Science
The simple dynamics of super Turing theories

Theoretical Computer Science - Special issue on universal machines and computations
Dynamical recognizers: real time language recognition by analog computers

Theoretical Computer Science
Neural networks and analog computation: beyond the Turing limit

Neural networks and analog computation: beyond the Turing limit
Introduction To Automata Theory, Languages, And Computation

Introduction To Automata Theory, Languages, And Computation
Simple Recurrent Networks Learn Context-Free and Context-Sensitive Languages by Counting

Neural Computation
Learning exponential state-growth languages by hill climbing

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

We describe a computational framework for language learning and parsing in which dynamical systems navigate on fractal sets. We explore the predictions of the framework in an artificial grammar task in which humans and recurrent neural networks are trained on a language with recursive structure. The results provide evidence for the claim of the dynamical systems models that grammatical systems continuously metamorphose during learning. The present perspective permits structural comparison between the recursive representations in symbolic and neural network models.