Machine Translation Comparison - Phase 2

Phase 2 of our Machine Translation comparison research is meant to build upon and elaborate our findings from Phase 1 . During Phase 1, we observed that for phrases under 140 characters, each translation engine began to exhibit clear strengths for translation between certain pairs of languages. For example, under 140 characters, Google excelled in French and Bing excelled in German, however these effects were not as pronounced for longer phrases (click here to see results).

In order to explore this effect in greater detail we have built Phase 2 to not only focus on shorter phrases but also include a great range of engines, adding the latest engines from Systran, Babylon, and Promt to be compared against Google Translate, Bing Translator and Yahoo Babelfish.

In addition, the new "Compare Translations" engine has been improved to allow users to auto generate sentences for comparison. This feature allows the voters to more easily compare participate in our research and also givens us a chance to see how multiple users vote on the same set of data.

We will continue to work to gather data and publish new results based on our ongoing research. Please participate and as always, thank you for taking interest and supporting our project.

Respectfully,

Gabble On Research Team