Discussion

• It is not very easy to quantify the contribution of each data set to the performance of Machine Translation systems. Some slides presented during ELRC Workshops were related to this issue (see in particular slide 63 here: http://www.lr-coordination.eu/sites/default/files/presentation/GER_workshop-master-S7b-Khalid%20Choukri-EN%20new.pdf)
  
  A paper written by Dragos Stefan Munteanu and Daniel Marcu, Information Sciences Institute, University of Southern California, was published on this subject and can be found here: http://www.aclweb.org/anthology/J05-4003

  

  • by The ELRC Helpdesk Team
  • at 5 years, 9 months ago