Search

This chapter aims to illustrate the potential of the dynamic usage-based (DUB) approach for yielding insights into the interrelationships between various complexity dimensions in second language (L2) learners’ performance and proficiency in the course of L2 development by applying a number of techniques proposed in the DUB literature to the L2 English writing productions of 10 Dutch-speaking secondary school learners over a period of two school years. The results reveal a high degree of variability, both within and across learners, and show that different complexity dimensions do not develop in parallel and may even enter into competitive relationships. Our findings demonstrate the importance of including multiple complexity dimensions in our analyses of L2 development in order to obtain a full and comprehensive picture of the development process. They also underline that one must be cautious both in projecting group findings onto individual learners as well as in generalising from individual case studies.

When using computer-aided translation systems in a typical, professional translation workflow, there are several stages at which there is room for improvement. The SCATE (Smart Computer-Aided Translation Environment) project investigated several of these aspects, both from a human-computer interaction point of view, as well as from a purely technological side. This paper describes the SCATE research with respect to improved fuzzy matching, parallel treebanks, the integration of translation memories with machine translation, quality estimation, terminology extraction from comparable texts, the use of speech recognition in the translation process, and human computer interaction and interface design for the professional translation environment. For each of these topics, we describe the experiments we performed and the conclusions drawn, providing an overview of the highlights of the entire SCATE project. The research in this project was funded by the Flemish Agency for Innovation and Technology IWT, project number 13007.

We identify a number of aspects that can boost the performance of Neural Fuzzy Repair (NFR), an easy-to-implement method to integrate translation memory matches and neural machine translation (NMT).We explore various ways of maximising the added value of retrieved matches within the NFR paradigm for eight language combinations, using Transformer NMT systems. In particular, we test the impact of different fuzzy matching techniques, sub-word-level segmentation methods and alignment-based features on overall translation quality. Furthermore, we propose a fuzzy match combination technique that aims to maximise the coverage of source words. This is supplemented with an analysis of how translation quality is affected by input sentence length and fuzzy match score. The results show that applying a combination of the tested modifications leads to a significant increase in estimated translation quality over all baselines for all language combinations. ispartof: Informatics vol:8 issue:1 pages:1-27 status: published

We present a simple yet powerful data augmentation method for boosting Neural Machine Translation (NMT) performance by leveraging information retrieved from a Translation Memory (TM). We propose and test two methods for augmenting NMT training data with fuzzy TM matches. Tests on the DGT-TM data set for two language pairs show consistent and substantial improvements over a range of baseline systems. The results suggest that this method is promising for any translation environment in which a sizeable TM is available and a certain amount of repetition across translations is to be expected, especially considering its ease of implementation.

This study aims to ascertain the nature and extent of the development of English L2 writing proficiency of 45 adult ESL learners over the time of an intensive short-term academic English language programme by means of quantitative measures targeting different components of the lexical and syntactic complexity of the learners’ writing performance, and to compare the scores on these measures with subjective ratings of learners’ overall writing quality. Results reveal several linguistic complexity measures that can adequately and validly capture changes in L2 writing in short-term ESL courses, though these do not include “popular” measures such as subordination ratios and lexical richness/diversity measures. Results also suggest that different subcomponents of syntactic and lexical complexity in L2 writing develop at a different pace, underlining the importance of calculating a sufficiently wide range of judiciously selected complexity measures in order to get a comprehensive picture of L2 writing development. Interestingly, the set of progress-sensitive complexity measures identified in this study does not coincide with the set of complexity measures that best predict subjective perceptions of writing quality.

This paper reports on a study on the nature and extent of the development of English L2 writing proficiency of 45 adult ESL learners over the time of an intensive short-term EAP program as evaluated by means of objective measures targeting different components of lexical and syntactic complexity. In addition, we compare the scores on these measures with more holistic and subjective ratings of learners' overall writing quality. Results reveal that some measures, but not necessarily the most popular linguistic complexity measures (e.g., subordination ratios and lexical richness measures), can indeed adequately and validly capture development in L2 writing in short-term ESL courses. Results further suggest that different subcomponents of syntactic and lexical complexity in L2 writing develop at different rates, which stressed the importance of calculating a sufficiently wide range of complexity measures in order to obtain a comprehensive picture of L2 development.