Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

PatPho: a phonological pattern generator for neural networks.

Ping Li1, Brian MacWhinney

  • 1Department of Psychology, University of Richmond, Richmond, VA 23173, USA. pli@richmond.edu

Behavior Research Methods, Instruments, & Computers : a Journal of the Psychonomic Society, Inc
|October 25, 2002
PubMed
Summary

A new tool, PatPho, generates accurate phonological representations for large English lexicons. This method effectively captures word similarities for neural network language models.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

The collection of speech data for the assessment of cognition remotely: Balancing ethical and practical challenges.

Alzheimer's & dementia (Amsterdam, Netherlands)·2026
Same author

Developmental Sentence Scoring for Preschool Language Sample Analysis: A Psychometric Update.

American journal of speech-language pathology·2026
Same author

Introducing the NEURAL Research Lab Data Set for Studies of Discourse and Gesture in Aphasia and Cognitively Healthy Aging Adults.

Journal of speech, language, and hearing research : JSLHR·2025
Same author

Priorities for New Data Collection.

Developmental science·2025
Same author

Mental simulation in bilingual and second language processing: New directions in the Competition Model.

Brain and language·2025
Same author

Six advances in research on bilingualism.

Brain and language·2025

Area of Science:

  • Computational Linguistics
  • Cognitive Science
  • Artificial Intelligence

Background:

  • Neural network models for language rely on statistical patterns in word phonology.
  • Existing methods for phonological representation struggle with large lexicons or accurate word depiction.

Purpose of the Study:

  • To introduce PatPho, a novel phonological pattern generator.
  • To enable connectionist modelers to obtain precise phonological representations of the English lexicon.

Main Methods:

  • Development of the PatPho tool for generating phonological patterns.
  • Testing PatPho's scalability with realistically sized lexicons.
  • Evaluating PatPho's accuracy in capturing phonological similarity structures.

Related Experiment Videos

Main Results:

  • PatPho successfully generates phonological patterns for large-scale English lexicons.
  • The tool accurately represents both monosyllabic and multisyllabic words.
  • PatPho efficiently captures phonological similarities.

Conclusions:

  • PatPho offers an effective solution for phonological representation in neural network language modeling.
  • The tool enhances the ability to model language use and acquisition using statistical phonological properties.