Using Hand-Written Rewrite Rules to Induce
            Underlying Morphology
                                     Michael A. Tepper
                                  University of Washington
                                 mtepper@u.washington.edu


                                           Abstract
     Allomorphic variation, or form-variation among morphemes with the same referential
     meaning, is often mentioned as a stumbling block to unsupervised morphological in-
     duction. To address this problem head-on, we present a hybrid approach that uses
     a small amount of linguistic knowledge in the form of orthographic rewrite rules to
     help refine an existing segmentation. Our goal is to learn when surface morphs (units
     of the segmentation) should really be counted together as the same underlying mor-
     pheme. In order to do this, we customize the Morfessor algorithm and model developed
     by Mathias Creutz and Krista Lagus, adding segmentation analyses generated by or-
     thographic rewrite rules along with a statistical framework to predict when analyses
     should be used as underlying morphemes. An initial segmentation produced by Morfes-
     sor Categories-MAP 0.9.2 is used as input. To suggest underlying morphemes, a set of
     language-specific orthographic rules is currently needed. Though we are not officially
     a part of the Challenge competition, for English and Turkish we report 62.22% and
     54.83% contest F-measures, which amount to 2% and 48% improvements respectively
     over top unsupervised entrants for those languages.

Keywords
Morphological induction, Allomorphic variation, Knowledge-lite, Word segmentation