Carolina Scarton has worked on the Framework for QE: QuEst (http://www.quest.dcs.shef.ac.uk/) - new version called QuEst++ with support for word, sentence and document-level QE (feature extraction and prediction). More specifically her contribution was on re-factoring core functions and features implementation for document-level QE. 

The direct link for the code is https://github.com/ghpaetzold/questplusplus and more information can be found in

Lucia Specia, Gustavo Henrique Paetzold and Carolina Scarton (2015): Multi-level Translation Quality Prediction with QuEst++. In the Proceedings of ACL-IJCNLP 2015 System Demonstrations, Beijing, China, pp. 115-120. http://www.aclweb.org/anthology/P15-4020.pdf