Learning Transducer Models for Morphological Analysis from Example Inflections

摘要

In this paper, we present a method to convert morphological inflection tables into unweighted and weighted finite transducers that perform parsing and generation. These transducers model the inflectional behavior of morphological paradigms induced from examples and can map inflected forms of previously unseen word forms into their lemmas and give morphosyntactic descriptions of them. The system is evaluated on several languages with data collected from the Wiktionary.