Skip to main content

As a part of the series "Translating into the Future" the Federal Association of Translators and Interpreters (Bundesverband der Dolmetscher und Übersetzer e.V.) will organize its next international conference with the motto "Translating and Interpreting 4.0 – New ways in the Digital Age". In addition to presentations, short seminars and workshops, participants may also take the possibility to meet old and new faces.

Daniel Zielinski and some of his collegues will also hold two presentations:

Terminology problems in machine translation

Machine translation (MT) has been on the rise for several years again. The development of neural machine translation considerably improved the quality of machine translation output. In comparison to former machine translations, the current outputs became more fluent for many language pairs and the number of grammatical errors has fallen significantly in some cases. A category which is still known to be a problem is terminology. While these errors may still be acceptable in raw machine translation for information purposes, they represent a problem in professional (technical) translation. The consistent usage of preferred terminology cannot be guaranteed and often resembles a Russian roulette game. In this seminar, we first explain how neural machine translation systems work. We will focus on how such systems are created and trained in order to explain the decision for one term or the other. For this, we will compare approaches of leading providers. Then, we will show the possibilities offered by various systems in order to improve the translation in the field of terminology. We will demonstrate this by using selected case studies. Last but not least, we look at further measures that can be implemented into the translation process in order to ensure a consistent usage of terminology. We will mainly focus on the use of automated terminology checking procedures. Using practical examples, we show the possibilities and limitations of checking mechanisms integrated in CAT tools.

Artificial intelligence during translation

In this workshop, we will compare the functionality as well as specific strengths and weaknesses of human and artificial intelligence. We will discuss how these may optimally complement each other during different phases of the translation process. Using practical examples, the importance of human language knowledge and technical expertise as well as the preparation for artificial intelligence (data preparation, creation of ontologies, training of engines, etc.) will be discussed. The central question is which aspects of human knowledge can be made accessible to machines – and which not. At the same time, this provides an outlook on the forthcoming changes in the profession of translators. Which areas of the translation process can be covered by machine translation in future (or already today) and why? How will man and machine work together and what skills does a translator require to do so? What are the requirements for clients related to knowledge, quality and risk management? Answers to these and other questions will be worked out during the workshop and presented by means of flip charts which can be supplemented by the participants. The participants of the workshop will also receive a handout with information on the topics covered.

You are cordially invited to attend the workshops and participate in the discussions.

Meet us

Meet us at BDÜ conference 2019 in Bonn and get to know us! If you would like to arrange a meeting with us, let us know: Send an email to info@loctimize.com or call us at + 49 (681) 91038365.

Additional Information

You can find more details about the BDÜ conference here