Understanding the Lexical Simplification Needs of Non-Native Speakers of English

Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, pages 717–727, Osaka, Japan, December 11-17 2016. Understanding the Lexical Simplification Needs of Non-Native Speakers of English Gustavo Henrique Paetzold and Lucia Specia Department of Computer Science University of Sheffield, UK {g.h.paetzold,l.specia}@sheffield.ac.uk Abstract We report three user studies in which the Lexical Simplification needs of non-native English speakers are investigated. Our analyses feature valuable new insight on the relationship between the non-natives’ notion of complexity and various morphological, semantic and lexical word properties. Some of our findings contradict long-standing misconceptions about word simplicity. The data produced in our studies consists of 211,564 annotations made by 1,100 volunteers, which we hope will guide forthcoming research on Text Simplification for non-native speakers of English. 1 Introduction Text Simplification is a useful application both to improve other Natural Language Processing tasks and to assist language-impaired readers (Chandrasekar et al., 1996). When a simplifier aims to help people, understanding their needs becomes very important. In Lexical Simplification – the task of replacing complex words and expressions with simpler alternatives – this has been shown to be the case (Rello et al., 2013b; Rello et al., 2013a; Rello et al., 2013c). They describe several user studies conducted with readers suffering from Dyslexia and outline the most recurring challenges faced by them, as well as the most effective ways to overcome these challenges. Given the widespread availability of content in English, non-native speakers of English become an important group to focus on. Reves and Medgyes (1994) elaborate on the differences between native and non-native English teachers from an educational and behavioural standpoint. However, we were not able to not find user studies that investigate the needs of non-native speakers from the perspective of Text Simplification. In this paper, we introduce three user studies that aim to do so. Each user study pertains to one of three steps in the usual Lexical Simplification pipeline (Figure 1): Complex Word Identification, Substitution Selection and Substitution Ranking. In the sections that fol- low, we describe the findings of each user study. Figure 1: Common Lexical Simplification Pipeline This work is licenced under a Creative Commons Attribution 4.0 International Licence. Licence details: http:// creativecommons.org/licenses/by/4.0/ 717 2 Complex Word Identification Complex Word Identification (CWI) is the task of deciding which words should be simplified in a text. Effective CWI strategies identify words which should not be simplified, and hence prevent LS systems from making inappropriate replacements (Paetzold, 2015). As shown in (Paetzold and Specia, 2013; Shardlow, 2014), ignoring CWI can considerably decrease the quality of the output produced by a sim- plifier. The goals of our user study on CWI are to: • Provide a better understanding on features of words that challenge non-native English speakers, and • Create a dataset that allows us to conceive models that automatically identify complex words. 2.1 Data Sources We selected 9,200 sentences with 20-40 words in length, at random, from three sources: • CW Corpus (Shardlow, 2013b): Composed of 731 sentences from …

Content truncated to 500 words