Morfessor EM+Prune: Improved subword segmentation with expectation maximization and pruning

Stig Arne Grönroos, Sami Virpioja, Mikko Kurimo

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

2 Citations (Scopus)
7 Downloads (Pure)

Abstract

Data-driven segmentation of words into subword units has been used in various natural language processing applications such as automatic speech recognition and statistical machine translation for almost 20 years. Recently it has became more widely adopted, as models based on deep neural networks often benefit from subword units even for morphologically simpler languages. In this paper, we discuss and compare training algorithms for a unigram subword model, based on the Expectation Maximization algorithm and lexicon pruning. Using English, Finnish, North Sami, and Turkish data sets, we show that this approach is able to find better solutions to the optimization problem defined by the Morfessor Baseline model than its original recursive training algorithm. The improved optimization also leads to higher morphological segmentation accuracy when compared to a linguistic gold standard. We publish implementations of the new algorithms in the widely-used Morfessor software package.

Original languageEnglish
Title of host publicationProceedings of The 12th Language Resources and Evaluation Conference
EditorsNicoletta Calzolari, Frederic Bechet, Philippe Blache, Khalid Choukri, Christopher Cieri, Thierry Declerck, Sara Goggi, Hitoshi Isahara, Bente Maegaard, Joseph Mariani, Helene Mazo, Asuncion Moreno, Jan Odijk, Stelios Piperidis
PublisherEuropean Language Resources Association (ELRA)
Pages3944-3953
Number of pages10
ISBN (Electronic)9791095546344
Publication statusPublished - 2020
MoE publication typeA4 Article in a conference publication
EventInternational Conference on Language Resources and Evaluation - Marseille, France
Duration: 11 May 202016 May 2020
Conference number: 12

Conference

ConferenceInternational Conference on Language Resources and Evaluation
Abbreviated titleLREC
Country/TerritoryFrance
CityMarseille
Period11/05/202016/05/2020

Keywords

  • Language Modelling
  • Less-Resourced/Endangered Languages
  • Morphology
  • Statistical and Machine Learning Methods
  • Tools
  • Unsupervised learning

Fingerprint

Dive into the research topics of 'Morfessor EM+Prune: Improved subword segmentation with expectation maximization and pruning'. Together they form a unique fingerprint.

Cite this