What is it about?
We produce a system which can automatically predict the quality of machine translations. The system can compare the output by different systems. This paper takes advantage of new quality indicators by using advanced statistical methods.
Featured Image
Why is it important?
We advance the state-of-the art in comparative quality estimation by suggesting new qualitative features and advanced ensemble methods for machine learning. We test the findings on the output of various Machine Translation systems produced during 7 years of progressive Machine Translation research and development
Perspectives
This paper gives a good perspectives on the indications one can use for comparing different MT outputs. It gives an empirical answer to questions such as "what are the most common errors and which of them are important for comparing MT outputs?"
Mr Eleftherios Avramidis
German Research Center for Artificial Intelligence (DFKI)
Read the Original
This page is a summary of: Comparative Quality Estimation for Machine Translation Observations on Machine Learning and Features, Prague Bulletin of Mathematical Linguistics, January 2017, De Gruyter,
DOI: 10.1515/pralin-2017-0029.
You can read the full text:
Contributors
The following have contributed to this page
