Collecting, Tagging, And Annotating Intersemiotic Mismatches Generated By Machine Translation

Authors

  • Augusto Espindola Universidade de Brasília
  • Thiago Pires Universidade de Brasília

Abstract

Technological advances have raised interest in multimodal corpus research. The development of tools designed for data collection, tagging, and annotation shows increasing progression. Although the profusion of such tools is promising, research aimed at text-image multimodal corpora is still incipient. Given that the variety of tools available makes it time-consuming for researchers to choose, this study proposes an analysis of tools that may assist collecting, tagging, and annotating processes of text-image multimodal documents in a multidisciplinary approach involving multimodality, machine translation and intersemiosis. To this end, this investigation is based on an adaptation of the findings in Rohlfing et al. (2006) and in the studies of intersemiotic mismatches generated by machine translation in Pires (2017). The findings point out that non-specialized tools have greater potential to make the processes of static multimodal documents annotation and tagging less time-consuming than the specialized tools.

Author Biography

  • Thiago Pires, Universidade de Brasília
    Doutor em Gestão da Informação pela Universidade de Brasília; Professor do Departamento de Línguas Estrangeiras e Tradução, vinculado ao Instituto de Letras da Universidade de Brasília.

Published

28/04/2020

How to Cite

Espindola, A., & Pires, T. (2020). Collecting, Tagging, And Annotating Intersemiotic Mismatches Generated By Machine Translation. Cultura E Tradução, 6(1), 248-264. https://periodicos.ufpb.br/index.php/ct/article/view/48886