Institute for Interpreting and Translation Studies at Stockholm University
About the session
Human vs. Machine – what machine translation needs to get right in subtitling
As the need for both interlingual and intralingual subtitling increases dramatically, many companies and authorities have started using methods such as speech recognition and machine translation to increase productivity and curb costs. These practices have led to an increase in quantity, but what about quality? We have compared traditional interlingual subtitles with subtitles created using machine translation and human postediting, to find out how quality differs and which features machine translated subtitles have. The results show that that the quality of machine translated subtitles is significantly lower, and display features that may interfere with audience satisfaction We want to point out some aspects where there is room for improvement, and also make a few suggestions for change.
Jan Pedersen is Associate Professor in Translation Studies and Director of the Institute for Interpreting and Translation Studies at Stockholm University, where he also researches and teaches audiovisual translation. He is a frequent presenter at international conferences, and his publications include the 2011 monograph Subtitling Norms for Television, as well as many articles on subtitling, translation and linguistics. He also worked as a television subtitler for many years