HAITrans - Human and Artificial Intelligence in Translation

HAITrans - Human and Artificial Intelligence in Translation - is a research group based in the University of Vienna Centre for Translation Studies. It investigates the behavioural and cognitive effects which technologies such as machine translation and automatic speech recognition and synthesis have on translators, as well as their impact on the profession, practice, training and society at large.

At present, the core research areas of the Vienna HAITrans Group are:

  1. Effects of speech technologies (Speech to Text and Text to Speech) on translation, revision and post-editing machine translation (PEMT) tasks (R1)
  2. Technology-supported translation, revision, and PEMT practices (R2)
  3. (Translation) technology for accessibility (R3)
  4. Translation technology didactics (R4)

In our qualitative and quantitative investigations we use data gathered via eye-tracking, questionnaires, focus groups, corpora, and translation environment tool metrics. We also collaborate with academic partners, international organisations, language service providers, dedicated professional associations and cultural-sector partners. 

Call for participation!

We are looking for professional translators based in or around Vienna, Austria, working from English into German who have German as a first language to contribute from the second half of January, 2023, to the research project “The impact of Speech Synthesis on cognitive load, productivity and quality during post-editing machine translation (PEMT)”. The project is led by Univ.-Prof. Dragoș Ciobanu and funded by Imminent (the research arm of Translated).

HAITrans is committed to researching novel uses of new and existing technologies which help translators adapt to the changing nature of the translation market, deliver high-quality translations, and improve productivity. The current research project will investigate the impact which speech synthesis (i.e. hearing an artificial, computer-generated voice reading the source and target segments) has on translators who post-edit machine translation output from English into German.

Your participation will consist of answering one pre-experiment and two experiment online questionnaires, as well as of taking part in an eye-tracking study in our Vienna-based lab where your eye movements on a computer screen, the computer sound, as well as your keyboard presses and mouse clicks will be recorded. Your face and voice will not be recorded. Your data will be fully anonymised within one month of your participation, so it will not be possible to trace it back to you.

Your task in this experiment will be to post-edit four short documents (factsheets about COVID-19 with a combined total number of words of 1,423 for the English original) in the Matecat CAT tool project, at times hearing the source and target segments being spoken out, and at other times without hearing any such synthetic sound.

An eye tracker will be used to record your gaze during the experiment. The computer screen and computer interactions you will have during the experiment will be recorded for later annotation and anonymous comparison with other experiment participants. Your details will be anonymised and you will be assigned a random experiment ID, so it will not be possible to trace back your performance to you personally. The effectiveness of your post-edits will be evaluated separately and will also be anonymised.

The total duration of the experiment will be up to three hours and you will be paid 100 EUR for your participation. You will need to fill in a University of Vienna freelancer datasheet and invoice the University of Vienna Centre for Translation Studies. You will be responsible for paying any tax associated with this amount as part of your earnings.

To register your interest, please fill in our recruitment questionnaire in English or in German. We will contact you shortly afterwards if your profile matches our experiment requirements.