Machine voice-over translation in social media: Perceived quality of translation of YouTube travel vlogs

Abstract

One of the challenges of social media as a means of interaction is to ensure media accessibility, i. e. to provide users with a free and unhindered access to the content of computer-mediated communication platforms. In the context of global networking, linguistic constraints due to the use of a verbal code different from the language of the potential user while creating media content, are not the least of the obstacles to such access. Given the obvious dominance of English in the online space, this is mainly a matter of expanding the audience by engaging individuals whose source language proficiency is not sufficient for unimpeded use of social media, in particular for viewing, listening and understanding video content uploaded on video hosting sites and other digital platforms. Since user-generated video content is generally characterised by individual and independent production, free accessibility and target audiences sharing a common interest for a specific subject matter, regardless of their residence area and language affiliation, human audio-visual translation is not considerable in such circumstances. For this reason, as a way to overcome language barriers and ensure media accessibility, the most advanced Internet portals are beginning to implement Web-based online machine translation solutions, made possible by the development of linguistic digital technologies. The survey conducted as a part of this study has shed light on the users’ perception of machine voice-over translation of English-language travel vlogs into Russian, and formulated some hypotheses regarding the ways to further improve its quality. The empirical evidence for this study was obtained from a two-step survey involving users and experts (audiovisual translators) who assessed the quality of machine voice translation of English-language travel vlog into Russian. According to the results ontained machine translation seems to be conceptually acceptable as a means of enabling language media accessibility in social media.