MT in Translation - A Holistic Approach
- Wednesday 1 December 2021, 12pm
- Zoom-link will be sent to network subscribers before the session
Machine learning (ML) and machine translation (MT) technologies have increasingly become involved in translation studies and practices. Compared with human's intuitive reflections on language, machine perception is largely based on mathematical calculation and statistical inference. While a neural machine translation engine aims to mimic human’s cognitive behavior, it is a learning machine in nature and thus is able to create its own subconscious area. This talk will take a holistic approach to translation, with MT as a component of the traditional concept of translation. It compares the results of English-Chinese translation that are created by human translators and machine translators using automatic MT evaluation tools.
The talk aims to generate discussions about the impact of MT on the teaching of translation. Some questions that can serve as a starting point for the discussion are:
- Is the concept of translation influenced by the rapid development of MT & ML? If so, how?
- Will the use of MT and other ML technologies change the cognitive aspects of human translators? If so, do you see a more positive or negative influence of MT on their minds?
- Will there be a relatively clear division of labor between human translators and machine translators in the future? If so, what do you think it will look like?
Keywords: machine translation (MT), machine learning (ML), statistical inference
Speaker bio: Dr. Peng Wang has taught, researched and practised translation and localization on three continents. She is the convener for EDUinLOC, a part-time professor at the University of Ottawa, and a member of the Conference Interpretation Advisory Panel for the Translation Bureau of the Canadian government. She is an instructor at the Localization Institute and her machine translation master class provides an independent perspective on MT implementation as well as its impact on the whole localization ecosystem. Her current research interests include human learning vs. machine learning, machine translation risk management, terminology and multilingual data analysis.