| Summary: | Computers have been used for natural language translation sincethe 1940s. The role for computers to support translation work has expandedsince the commercialisation of the Internet in the early 1990s.As an implication of ’crowd sourcing’, translation is now supportedby users through community translation, meaning that people - usersof web sites - actively translate web sites into different languages. Inthis thesis we will present Babbler - a novel concept for communitytranslation. Following a design science research approach, a softwarewas designed and implemented in an information systems developmentproject. We present our conceptual design and its software implementation,and evaluate it using different techniques, including log analysis,interviews with translators, and an informed argument contrastingour design to other community translation approaches. The evaluationaddresses various qualities of the design, including effectiveness, efficiency,reliability, workflow, implementability and performance. Basedon our results, we reflect about translations to different target groups(based on demographics such as gender, age and culture), informedby an explorative analysis of translation results in the empirical context.We also show further implications for future design of communitytranslation artefacts, and future research in the area.
|
|---|
