An Investigation of Warning Erroneous Chat Translations in Cross-lingual Communication

TL;DR

How individuals perceive warning messages about potential mistranslations and whether they benefit the crowd is investigated to investigate and warning messages' contribution to making chat translation systems effective is demonstrated.

Abstract

Machine translation models are still inappropriate for translating chats, despite the popularity of translation software and plug-in applications. The complexity of dialogues poses significant challenges and can hinder crosslingual communication. Instead of pursuing a flawless translation system, a more practical approach would be to issue warning messages about potential mistranslations to reduce confusion. However, it is still unclear how individuals perceive these warning messages and whether they benefit the crowd. This paper tackles to investigate this question and demonstrates the warning messages' contribution to making chat translation systems effective.

Figures (5)

• Figure 1: An illustration of the designed survey. Participants will engage in two rounds of chat in the survey: one without warning messages (left) and one with warning messages (right). The content and response options are the same in both rounds. The order of the two rounds, either "without-with" (solid line) or "with-without" (dotted line), will be randomly assigned to participants.
• Figure 2: The responses to how participants think the warning messages helped them continue the chat.
• Figure 3: The results that whether participants changed their choices with the help of warning messages.
• Figure 4: The responses to how participants think the warnings of the received/sent messages helped them continue the chat.
• Figure 5: The results about expected additional features to the warning messages.