The Hebrew Essay Corpus

Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 5580–5586 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 5580 The Hebrew Essay Corpus Chen Gafni, Anat Prior, Shuly Wintner University of Haifa, Israel chen.gafni@gmail.com, aprior@edu.haifa.ac.il, shuly@cs.haifa.ac.il Abstract We present the Hebrew Essay Corpus: an annotated corpus of Hebrew language argumentative essays authored by prospective higher-education students. The corpus includes both essays by native speakers, written as part of the psychometric exam that is used to assess their future success in academic studies; and essays authored by non-native speakers, with three different native languages, that were written as part of a language aptitude test. The corpus is uniformly encoded and stored. The non- native essays were annotated with target hypotheses whose main goal is to make the texts amenable to automatic processing (morphological and syntactic analysis). The corpus is available for academic purposes upon request. We describe the corpus and the error correction and annotation schemes used in its analysis. In addition to introducing this new resource, we discuss the challenges of identifying and analyzing non-native language use in general, and propose various ways for dealing with these challenges. Keywords: Learner corpora, Hebrew, non-native language 1. Introduction Learner corpora—the systematic collection of spo- ken or written language produced by learners of a language—have been used in research since the late 1980’s (Granger, 2002; Tono, 2003; Granger et al., 2015; De Knop and Meunier, 2015). They can fol- low different designs, be of different sizes, involve different language pairs, etc.1 One paradigm in an- alyzing learner corpora is the quantitative compari- son of categories (words, multi-word expressions, parts of speech, etc.) between learner corpora and na- tive speaker corpora (Granger, 1996; Gilquin, 2008; Granger, 2015). This approach, which we follow here, is often called Contrastive Interlanguage Analysis. The quantitative analyses range from descriptive compar- isons (overuse/underuse studies (Durrant and Schmitt, 2009; Gilquin and Paquot, 2008; Hirschmann et al., 2013)) to more involved statistical methods, up to mod- eling (Gries, 2008; Gries, 2015; Gries and Deshors, 2015; Vyatkina et al., 2015). Learner and other non-native language corpora have been shown to be instrumental in several tasks, includ- ing automatic detection of highly competent non-native writers (Tomokiyo and Jones, 2001; Estival et al., 2007; Bergsma et al., 2012), identification of learners’ na- tive language (Koppel et al., 2005; Bykh and Meur- ers, 2012; Tetreault et al., 2013; Tsvetkov et al., 2013; Goldin et al., 2018) and typology-driven error predic- tion in learners’ language production (Berzak et al., 2015). In this paper we present the Hebrew Essay Corpus: an annotated corpus of Hebrew language argumenta- tive essays authored by prospective students in higher- education. The corpus includes both essays by native (or near-native) speakers, written as part of a college 1For a list of learner corpora, see Learner Corpora around the World; for an extensive bibliography covering learner cor- pus analyses, see the resources page of the Learner Corpus Association. entry exam that is used to …

Content truncated to 500 words