Predicting reaction times in word recognition by unsupervised learning of morphology

Sami Virpioja, Minna Lehtonen, Annika Hulten, Riitta Salmelin, Krista Lagus

Forskningsoutput: Kapitel i bok/rapport/konferenshandlingKonferensbidragVetenskapligPeer review


A central question in the study of the mental lexicon is how morphologically complex words are processed. We consider this question from the viewpoint of statistical models of morphology. As an indicator of the mental processing cost in the brain, we use reaction times to words in a visual lexical decision task on Finnish nouns. Statistical correlation between a model and reaction times is employed as a goodness measure of the model. In particular, we study Morfessor, an unsupervised method for learning concatenative morphology. The results for a set of inflected and monomorphemic Finnish nouns reveal that the probabilities given by Morfessor, especially the Categories-MAP version, show considerably higher correlations to the reaction times than simple word statistics such as frequency, morphological family size, or length. These correlations are also higher than when any individual test subject is viewed as a model.
Titel på gästpublikationArtificial Neural Networks and Machine Learning - ICANN 2011 : 21th International Conference on Artificial Neural Networks, Espoo, Finland, June 14-17, 2011, Proceedings, Part I
RedaktörerTimo Honkela, Wlodzislaw Duch, Mark Girolami, Samuel Kaski
Antal sidor8
ISBN (tryckt)978-3-642-21734-0
StatusPublicerad - 2011
MoE-publikationstypA4 Artikel i en konferenspublikation
Evenemang21st International Conference on Artificial Neural Networks, ICANN 2011 - Espoo, Finland
Varaktighet: 14 jun 201117 jun 2011


NamnLecture Notes in Computer Science
ISSN (tryckt)0302-9743


  • 6121 Språkvetenskaper

Citera det här