Recursive Context-Aware Lexical Simplification

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, pages 4853–4863, Hong Kong, China, November 3–7, 2019. c⃝2019 Association for Computational Linguistics 4853 Recursive Context-Aware Lexical Simplification Sian Gooding Dept of Computer Science and Technology University of Cambridge shg36@cam.ac.uk Ekaterina Kochmar ALTA Institute Unversity of Cambridge ek358@cam.ac.uk Abstract This paper presents a novel architecture for recursive context-aware lexical simplification, REC-LS, that is capable of (1) making use of the wider context when detecting the words in need of simplification and suggesting alter- natives, and (2) taking previous simplification steps into account. We show that our system outputs lexical simplifications that are gram- matically correct and semantically appropri- ate, and outperforms the current state-of-the- art systems in lexical simplification. 1 Introduction Text simplification (TS) is aimed at reducing the reading and grammatical complexity of text while retaining the meaning and grammaticality (Chan- drasekar and Bangalore, 1997). This is usually achieved by a series of transformations at the lex- ical and syntactic level. A number of systems in the recent years have approached this task in an integral manner (Zhu et al., 2010; Kauchak, 2013; Zhang and Lapata, 2017). Such comprehensive systems can perform a number of simplification operations at once, but the results are sometimes ungrammatical and meaning can be changed, ar- guably making the original text less clear and more complex (Siddharthan, 2014). In this paper, we assume the lexical and syn- tactic components of a TS system to be indepen- dent and complementary to each other, and focus on the lexical simplification (LS) component for a number of reasons. First, it has been shown that lexical simplification techniques positively impact the readability of text and improve reader under- standing and information retention (Leroy et al., 2012). Secondly, it has been argued that a large number of people with reading difficulties, in- cluding those with disabilities, low-literacy, non- native backgrounds or non-expert knowledge ben- efit from LS (Xu et al., 2015). For instance, James (1998) shows that vocabulary of the non-native language plays central role in second language ac- quisition. As we aim to make information more accessible to such readers, the quality of the sim- plified text is of a paramount importance. We stress the importance of three key points of quality assessment in LS: it is vital for the simpli- fied text to be of a lower complexity, while being semantically equivalent to the original, and gram- matically correct. Context plays a central role in fulfilling these requirements. For instance, con- sider the different uses of situation in the complex word dataset collected by Yimam et al. (2017): (1) the gravity of the economic situation (2) the situation has remained unchanged This example demonstrates two types of contex- tual effects: first of all, in (1) both economic and situation are marked as complex in mutual con- text, but when situation occurs in a different con- text it is not annotated as complex. Secondly, this example …

Content truncated to 500 words