Merging Data Resources for Inflectional and Derivational Morphology in Czech

Merging Data Resources for Inflectional and Derivational Morphology in Czech Zdenˇek ˇZabokrtsk´y, Magda ˇSevˇc´ıkov´a, Milan Straka, Jon´aˇs Vidra, Ad´ela Limbursk´a Charles University in Prague, Faculty of Mathematics and Physics, Institute of Formal and Applied Linguistics {zabokrtsky,sevcikova,straka,vidra}@ufal.mff.cuni.cz, limburska@seznam.cz Abstract The paper deals with merging two complementary resources of morphological data previously existing for Czech, namely the inflectional dictionary MorfFlex CZ and the recently developed lexical network DeriNet. The MorfFlex CZ dictionary has been used by a morphological analyzer capable of analyzing/generating several million Czech word forms according to the rules of Czech inflection. The DeriNet network contains several hundred thousand Czech lemmas interconnected with links corresponding to derivational relations (relations between base words and words derived from them). After summarizing basic characteristics of both resources, the process of merging is described, focusing on both rather technical aspects (growth of the data, measuring the quality of newly added derivational relations) and linguistic issues (treating lexical homonymy and vowel/consonant alternations). The resulting resource contains 970 thousand lemmas connected with 715 thousand derivational relations and is publicly available on the web under the CC-BY-NC-SA license. The data were incorporated in the MorphoDiTa library version 2.0 (which provides morphological analysis, generation, tagging and lemmatization for Czech) and can be browsed and searched by two web tools (DeriNet Viewer and DeriNet Search tool). Keywords: derivation, inflection, morphology 1. Introduction The present paper deals with merging two complemen- tary resources of morphological data previously exist- ing for Czech: first, the inflectional dictionary MorfFlex CZ used by a morphological analyzer capable of analyz- ing/generating several million Czech word forms accord- ing to the rules of Czech inflection (Hajiˇc and Hlav´aˇcov´a, 2013), and second, the recently developed lexical net- work DeriNet that stores derivational relations among several hundred thousand Czech lemmas (ˇSevˇc´ıkov´a and ˇZabokrtsk´y, 2014). The resulting publicly available data re- source interconnects both types of information and thus al- lows exploiting the two morphological phenomena in NLP applications in a unified way. In addition, using the full in- flectional dictionary led to a considerable growth of both recall and precision of the captured derivational relations. 2. Related work Czech is a language with both a rich inflectional system and a complex derivational morphology. In the theoretical description, inflectional and derivational morphology have been traditionally separated from each other, derivations being described as a part of the word-formation system of Czech (in addition to compounding and combined word- formation processes); e.g. Dokulil et al. (1986), Karl´ık et al. (1995), ˇSt´ıcha (2013). The roots of the separa- tion are traced back by Bednaˇr´ıkov´a (2009), who states that in the representative, “academic” grammars of Czech (Dokulil et al., 1986; Kom´arek et al., 1986) “word forma- tion is seen as a sort of transition zone between morphol- ogy and lexicon” and “word formation means are consid- ered non-morphological means, along with lexical means, means of word order and intonation etc.” Although the linguistic tradition influenced also developers of morphological software tools for Czech, who mostly ap- proached inflection and derivation in different …

Content truncated to 500 words