This post contains two examples of projects I completed in my Advanced Translation Technology course at MIIS with colleagues Koni Chen, Diana Song, Baoze Zhang, Junyi Zhang, and Geyu Chen . The Machine Translation Project section contains the deliverables for our MT project, where we trained a machine translation engine for Education Tedtalks using Microsoft Custom Translator. The Translation Management Project section contains our video presentation of evaluating XTM and Wordbee and selecting the right TMS for company bambooHR’s needs. In this course, I learned about how translation technology can improve all aspects of the translation process, as well as how to train MT engines and navigate even the most complicated TMSs.
Machine Translation Project
Summary
Our MT project involved training an MT engine using Microsoft Custom Translator to translate Tedtalks on Education from English to Chinese. In the pilot proposal, we planned out the project, established quality, timing, and pricing goals, and estimated the timeline and associated costs. In the updated proposal, we amended our plan to reflect the knowledge we had gained and calculated the time and costs to fully train the engine with an updated timeline. Finally, in the presentation, we explained the information in our pilot and updated proposals, as well as any problems we faced and what we learned from the project. The most important lesson we learned is that the quality of date and proper segmentations were necessary and that training a machine would require larger costs and more time before reaching any desirable results. As for myself I took on the challenge of working with a language I did not know and learned how to catch errors in formatting and tags with an unknown language.
Pilot Proposal
Updated Pilot Project
Download updated deliverables here
Translation Management System Project
Summary
Our TMS project was a simulated comparison project, where we chose a client—in our case, bambooHR—and analyzed 2 translation management systems based on their suitability for the company’s workflow needs. We established a list of requirements for the TMSs, based on company’s needs. We then used this to create a scorecard from 1-10 in order to compare them, and weighed them based on whether they were must-have features or simply nice to have. Then, we evaluated XTM Cloud and Wordbee based on how well they would work for facilitating the completion of bambooHR requirements.
Presentation
