We assess the language specificity of recent language models by exploring the potential of a multilingual language model. In particular, we evaluate Google's multilingual BERT (mBERT) model on Named Entity Recognition (NER) in German and English. We expand the work on language model fine-tuning by Howard and Ruder (2018), applying it to the BERT architecture. We successfully reproduce the NER results published by Devlin et al. (2019). Our results show that the multilingual language model generalises well for NER in the chosen languages, matching the native model in English and comparing well with recent approaches for German. However, it does not benefit from the added finetuning methods.
This article sheds light on how conference interpreting practice has changed since the beginning of the pandemic in Austria by drawing on the sociological concepts “body” and “space” as analytical and conceptual categories, among others shaped by Lindemann and Lindemann and Schünemann. The results of this study yield insights into the different modes of spatial organisation for conference interpreting during the pandemic; the ways these different organisational settings impacted cooperation among interpreters; and the interpreters’ perceptions of these reconfigurations in space and bodily presence. The data, obtained through an online survey and interviews with in total nine conference interpreters in Austria, show that a range of organisational settings is possible when working (completely or partly) remote, all entailing their individual particularities. Among the observed settings were conferences held on-site with strict anti-COVID measures; “hybrid” settings where interpreters worked on-site with speakers joining online or vice versa; and remote interpreting assignments. In remote assignments, more communication was necessary before the conference, for example, to discuss handovers. Additional communication tools were resorted to, and interpreters preferred working together in one physical place, for example, to facilitate cooperation. However, they also saw advantages in remote assignments.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.