A methodology for the semiautomatic annotation of EPEC-RolSem, a Basque corpus labeled at predicate level following the PropBank-VerbNet model

Abstract

In this article we describe the methodology developed for the semiautomatic annotation of EPEC-RolSem, a Basque corpus labeled at predicate level that follows the PropBank-VerbNet model. The methodology presented is the product of detailed theoretical study of the semantic nature of verbs in Basque and of their similarities and differences with verbs in other languages. As part of the proposed methodology, we are creating a Basque lexicon on the PropBank-VerbNet model that we have named the Basque Verb Index (BVI). Our work thus dovetails with the general trend toward building lexicons from tagged corpora that is clear in work conducted for other languages. EPEC-RolSem and BVI are two important resources for the computational semantic processing of Basque; as far as the authors are aware, they are also the first resources of their kind developed for Basque. In addition, each entry in BVI is linked to the corresponding verb-entry in well-known resources like PropBank, VerbNet, WordNet, FrameNet, and Levin's classification. We have also implemented several automatic processes to aid in creating and annotating the BVI, including processes designed to facilitate the task of manual annotation.

Online resources for the syntactic-semantic classification of verbs

“…The Basque Verb Index (ESTARRONA et al, 2015) follows VerbNet's model and it is also focused on NLP applications. It can be accessed at http://ixa2.si.ehu.eus/e-rolda/ (accessed on: 24 Nov. 2021), together with other resources, using the e-ROLda tool.…”

“…Also, based on her initial ideas and inspired by VerbNet, many researchers around the world proposed similar classifications and resources for the verbal lexicon of languages other than English. BVI (ESTARRONA et al, 2015), AnCoraVerb (APARICIO; TAULÉ; MARTÍ, 2008; TAULÉ; MARTÍ; BORREGA, 2010), VerboWeb (CANÇADO; AMARAL; MEIRELLES, 2017), CROVALLEX (PRERADOVIC; BORAS; KISICEK, 2009;PRERADOVIC, 2010), and VALLEX (KETTNEROVÁ; LOPATKOVÁ; BEJČEK, 2012) are of particular interest in this paper, and they are all online resources, or digital lexicons, like VerbNet, for the analysis and syntactic-semantic classification of verbs in different languages (BVI for Basque, AnCoraVerb for Spanish and Catalan, VerboWeb for Brazilian Portuguese, CROVALLEX for Croatian, and VALLEX for Czech).…”

Verb Class Induction with Partial Supervision

“…This method uses a small amount of labeled data, in com-bination with state-of-the-art techniques for creating Levinstyle clusters, and is extremely practical. Our aim is to multiply the impact of annotation, and while our experiments focus on English, we believe this work can assist the creators of VerbNet-like resources in multiple languages (Mousser 2010;Danlos, Nakamura, and Pradet 2014;Pradet, Danlos, and De Chalendar 2014;Scarton, Duran, and Aluísio 2014;Estarrona et al 2015;Hautli-Janisz, King, and Ramchand 2015;Vulić, Mrkšić, and Korhonen 2017). Our main contribution is the demonstration that this simple supervision technique aids in the recovery of VerbNet classes.…”