Options
Paper

Online Learning of ITSL Grammars

Authors
  • Jacob K. Johnson (University of Utah)
  • Aniello De Santo (University of Utah)

Abstract

This paper presents the first incremental learning algorithm for input-sensitive TSL languages (ITSL). We leverage insights from De Santo and AksÎnova's (2021) ITSL batch-learner to extend Lambert's (2021) string extension learning approach to online learning of TSL. We discuss formal properties of the extension, and evaluate the effectiveness of both the original TSL learner and the new ITSL learner on a variety of phonotactic patterns, empirically demonstrating the effectiveness of both.

Keywords: subregular languages, incremental learning, TSL, ITSL, string extension learning, phonotactics

How to Cite:

Johnson, J. K. & De Santo, A., (2024) “Online Learning of ITSL Grammars”, Society for Computation in Linguistics 7(1), 257–267. doi: https://doi.org/10.7275/scil.2159

Downloads:
Download PDF

201 Views

69 Downloads

Published on
24 Jun 2024
Peer Reviewed
License
Creative Commons Attribution 4.0