Porting Multilingual Morphological Resources to OntoLex-Lemon

Proceedings of Recent Advances in Natural Language Processing, pages 233–238, Varna, Bulgaria, Sep 2–4, 2019. https://doi.org/10.26615/978-954-452-056-4_027 233 Porting Multilingual Morphological Resources to OntoLex-Lemon Thierry Declerck German Research Center for Artificial Intelligence Saarbrücken, Germany thierry.declerck@dfki.de Stefania Racioppa German Research Center for Artificial Intelligence Saarbrücken, Germany stefania.racioppa@dfki.de Abstract We describe work consisting in porting vari- ous morphological resources to the OntoLex- Lemon model. A main objective of this work is to offer a uniform representation of different morphological data sets in order to be able to compare and interlink multilingual resources and to cross-check and interlink or merge the content of morphological resources of one and the same language. The results of our work will be published on the Linguistic Linked Open Data cloud. 1 Introduction A significant number of linguistic resources have been published on the Linguistic Linked Open Data cloud (LLOD),1 and projects like ELEXIS2 and Prêt-à-LLOD3 are contributing to its further extension. But we notice that only very few, if any, specific morphological resources are included in the LLOD cloud. Available morphology informa- tion is mostly contained in lexical or dictionary en- tries. Our aim is to make also specialized morpho- logical resources available in this cloud. With this step we want to support the interlinking of such resources with other types of linguistic data, in a multilingual fashion, extending work described in (Gromann and Declerck, 2019), which is linking synsets of the Princeton WordNet (PWN) (Fell- baum, 1998) that are associated with plural forms to full lexical descriptions. A first step of our current work consisted in mapping the MMorph4 set of multilingual 1See http://www.linguistic-lod.org/ 2https://elex.is/. See also (Krek et al., 2018) for a general overview of ELEXIS and (Declerck et al., 2018) for a focused description of the role of LLOD in the context of the project. 3https://www.pret-a-llod.eu. 4See (Petitpierre and Russell, 1995). morphological resources to the OntoLex-Lemon model.5 In the next sections, we first describe briefly the Linguistic Linked Open Data cloud, and one of its salient component, the OntoLex-Lemon model. Following this summary, we describe the (multi- lingual) morphological resources we selected for mapping to OntoLex-Lemon. We present the re- sult of such mappings and conclude with a de- scription of the next steps of our work, aiming at supporting the cross-lingual comparison of mor- phological resources, and the cross-checking, cor- recting and merging of different morphological re- sources for one and the same language. 2 The Linguistic Linked Open Data Cloud The LLOD initiative had its inception in 2012 at a workshop co-located with the 34th Annual Con- ference of the German Linguistic Society (DGfS). The workshop was organized by members of the Open Knowledge Foundation,6 and the contribu- tions to this workshop are available in (Chiarcos et al., 2012). The workshop has been a point of focal activity for several research and infrastruc- ture projects, as well as for the “Ontology Lex- ica” W3C Community Group.7 Those develop- ments are described in (McCrae et al., 2016). A major result of those activities is …

Content truncated to 500 words