Controlling Text Complexity in Neural Machine Translation

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, pages 1549–1564, Hong Kong, China, November 3–7, 2019. c⃝2019 Association for Computational Linguistics 1549 Controlling Text Complexity in Neural Machine Translation Sweta Agrawal Department of Computer Science University of Maryland sweagraw@cs.umd.edu Marine Carpuat Department of Computer Science University of Maryland marine@cs.umd.edu Abstract This work introduces a machine translation task where the output is aimed at audiences of different levels of target language proficiency. We collect a high quality dataset of news arti- cles available in English and Spanish, written for diverse grade levels and propose a method to align segments across comparable bilingual articles. The resulting dataset makes it pos- sible to train multi-task sequence-to-sequence models that translate Spanish into English tar- geted at an easier reading grade level than the original Spanish. We show that these multi- task models outperform pipeline approaches that translate and simplify text independently. 1 Introduction Generating text at the right level of complexity can make machine translation (MT) more useful for a wide range of users. As Xu et al. (2015) note, sim- plifying text makes it possible to develop reading aids for people with low-literacy (Watanabe et al., 2009; De Belder and Moens, 2010), for non-native speakers and language learners (Petersen and Os- tendorf, 2007; Allen, 2009), for people who suffer from language impairments (Carroll et al., 1999; Canning et al., 2000; Inui et al., 2003), and for readers lacking expert knowledge of the topic dis- cussed (Elhadad and Sutaria, 2007; Siddharthan and Katsos, 2010). Such readers would also ben- efit from MT output that is better targeted to their needs by being easier to read than the original. Prior work on text complexity has focused on simplifying input text in one language, primar- ily English (Chandrasekar et al., 1996; Coster and Kauchak, 2011; Siddharthan, 2014; Xu et al., 2015; Zhang and Lapata, 2017; Scarton and Spe- cia, 2018; Kriz et al., 2019; Nishihara et al., 2019). Simplification has been used to improve MT by restructuring complex sentences into shorter and simpler segments that are easier to translate (Ger- ber and Hovy, 1998; Štajner and Popovic, 2016; Hasler et al., 2017). Contemporaneously to our work, Marchisio et al. (2019) show that tagging the English side of parallel corpora with automatic readability scores can help translate the same in- put in a simpler or more complex form. Our work shares the goal of controlling translation complex- ity, but considers a broader range of reading grade levels and simplification operations grounded in professionally edited text simplification corpora. Building a model for this task ideally re- quires rich annotation for evaluation and super- vised training that is not available in bilingual parallel corpora typically used in MT. Control- ling the complexity of Spanish-English translation ideally requires examples of Spanish sentences paired with several English translations that span a range of complexity levels. We collect such a dataset of English-Spanish segment pairs from the Newsela …

Content truncated to 500 words