Bridging the gap between machine translation output and images in multimodal documents

The aim of this article is to report on recent findings concerning the use Google Translate outputs in multimodal contexts. Development and evaluation machine translation often focus verbal mode, but accounts by area exploration text-image relations documents translated automatically are rare. Thus, work seeks describe just what such how them, organized two parts: firstly, exploring problem through an interdisciplinary interface, involving Machine Translation Multimodality analyze some examples from Wikihow website; secondly, reporting investigation suitable tools methods properly annotate these issues within a long-term purpose assemble corpus. Finally, provides discussion findings, including limitations perspectives for future research.