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

Unsupervised learning of natural languages.

Zach Solan1, David Horn, Eytan Ruppin

  • 1School of Physics and Astronomy, Tel Aviv University, Tel Aviv 69978, Israel.

Proceedings of the National Academy of Sciences of the United States of America
|August 10, 2005
PubMed
Summary

This study introduces an unsupervised algorithm that discovers hierarchical patterns in sequential data. The method aids in inferring underlying rules for linguistics and bioinformatics, enabling complex structure discovery.

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

A genome-wide genetic screen identifies a novel kDNA replication protein in trypanosomes.

Nucleic acids research·2026
Same author

Lossless resistive micro-heater design for reconfigurable phase-change photonics.

Optics letters·2026
Same author

FMOPhore for hotspot identification and efficient fragment-to-lead growth strategies.

Nature communications·2026
Same author

Genetic origins and proteomic consequences of kinetoplast loss in trypanosomes.

PLoS pathogens·2026
Same author

Acoziborole resistance associated mutations in Trypanosoma brucei CPSF3.

PLoS pathogens·2026
Same author

Decoding efficacy and resistance space at a drug binding site.

Nature communications·2026

Area of Science:

  • Computational linguistics
  • Bioinformatics
  • Machine learning
  • Pattern recognition

Background:

  • Inferring underlying rules from raw symbolic sequential data is a fundamental challenge.
  • Existing methods may struggle with complex, hierarchical structures inherent in diverse data types.

Purpose of the Study:

  • To develop an unsupervised algorithm for distilling hierarchical patterns from sequential data.
  • To enable the inference of governing rules in fields like linguistics and bioinformatics.

Main Methods:

  • The study presents the adios (automatic distillation of structure) algorithm.
  • This algorithm employs statistical pattern extraction and structured generalization.
  • It recursively distills hierarchical structures from input corpora.

Related Experiment Videos

Main Results:

  • The adios algorithm successfully learned complex syntax in artificial grammars and natural languages (English, Chinese).
  • It demonstrated capability in generating novel, grammatical sentences.
  • The algorithm showed utility in bioinformatics for correlating protein sequence with function.

Conclusions:

  • Unsupervised learning of hierarchical structures is feasible and effective.
  • The adios algorithm offers a powerful tool for structure discovery across various scientific domains.
  • This approach has significant implications for linguistics, bioinformatics, and beyond.