Interpretability for Language Learners Using Example-Based Grammatical Error Correction

Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers, pages 7176 - 7187 May 22-27, 2022 c⃝2022 Association for Computational Linguistics Interpretability for Language Learners Using Example-Based Grammatical Error Correction Masahiro Kaneko Sho Takase Ayana Niwa Naoaki Okazaki Tokyo Institute of Technology {masahiro.kaneko, sho.takase, ayana.niwa}@nlp.c.titech.ac.jp okazaki@c.titech.ac.jp Abstract Grammatical Error Correction (GEC) should focus not only on correction accuracy but also on the interpretability of the results for language learners. However, existing neural- based GEC models mostly focus on improv- ing accuracy, while their interpretability has not been explored. Example-based methods are promising for improving interpretability, which use similar retrieved examples to gen- erate corrections. Furthermore, examples are beneficial in language learning, helping learn- ers to understand the basis for grammatically incorrect/correct texts and improve their confi- dence in writing. Therefore, we hypothesized that incorporating an example-based method into GEC could improve interpretability and support language learners. In this study, we introduce an Example-Based GEC (EB-GEC) that presents examples to language learners as a basis for correction result. The examples con- sist of pairs of correct and incorrect sentences similar to a given input and its predicted cor- rection. Experiments demonstrate that the ex- amples presented by EB-GEC help language learners decide whether to accept or refuse sug- gestions from the GEC output. Furthermore, the experiments show that retrieved examples also improve the accuracy of corrections. 1 Introduction Grammatical Error Correction (GEC) models, which generate grammatically correct texts from grammatically incorrect texts, are useful for lan- guage learners. In GEC, various neural-based mod- els have been proposed to improve the correction accuracy (Yuan and Briscoe, 2016; Chollampatt and Ng, 2018; Junczys-Dowmunt et al., 2018; Zhao et al., 2019; Kaneko et al., 2020; Omelianchuk et al., 2020). However, the basis on which a neural GEC model makes corrections is generally uninter- pretable to learners. Neural GEC models rarely ad- dress correction interpretability, leaving language Figure 1: EB-GEC presents not only a correction but also an example of why the GEC model suggested this correction. learners with no explanation of the reason for a correction. Interpretability plays a key role in educational scenarios (Webb et al., 2020). In particular, present- ing examples is shown to be effective in improving understanding. Language learners acquire gram- matical rules and vocabulary from examples (Johns, 1994; Mizumoto and Chujo, 2015). Presenting ex- amples of incorrect sentences together with correct ones improves the understanding of grammatical correctness as well as essay quality (Arai et al., 2019, 2020). Recently, example-based methods have been ap- plied to a wide range of natural language process- ing tasks to improve the interpretability of neu- ral models, including machine translation (Khan- delwal et al., 2021), part-of-speech tagging (Wise- man and Stratos, 2019), and named entity recogni- tion (Ouchi et al., 2020). These methods predict labels or tokens by considering the nearest neigh- bor examples retrieved by the representations of the model at the inference time. Khandelwal et al. (2021) showed that in machine translation, exam- ples …

Content truncated to 500 words