The Impact of Automation and Metadata on Wikipedia: A Study of Translational Practices in Four Language Communities

Abstract

This discussion explores how different attitudes to machine translation (hereinafter MT) may lead to the perpetuation of certain translational practices. Drawing on fundamental concepts from practice theory (Wegner 1995, Nicolini 2012, Olohan 2017), this study assesses the importance of materials such as translation tools, software robots, and metadata for content-creation purposes in non-professional collaborative spaces like Wikipedia. More specifically, it delves into the implications of using software for translating articles from English into Spanish, Norwegian, Swedish, and Dutch within the user-driven platform Wikipedia. Both the Swedish and the Dutch communities of Wikipedia have adopted more flexible approaches about the use of MT, to the extent that at least 40% of the articles created in the Swedish Wikipedia in 2014 were produced by one single software robot. The other two communities, on the other hand, have either restricted or banned the use of these programs arguing that quality should prevail over quantity. Consequently, this work, which is part of large-scale project at the University of Manchester, examines the role of, and increasing reliance on, automation in translation. In doing so, special attention is paid to a) Wikipedia articles that have been created by robots and other tools, and b) how the four language communities approach MT in a different way, readjusting their policies accordingly. The ultimate goal is to have an interactive session with MT experts and enthusiasts.

Details

Presentation Type

Focused Discussion

Theme

2019 Special Focus - The World 4.0: Convergence of Knowledges and Machines

KEYWORDS

Wikipedia, Machine Translation, Automation, Translation Studies, Crowdsourcing

Digital Media

This presenter hasn’t added media.
Request media and follow this presentation.