AUTOMATIC PROCEDURES IN MT EVALUATION - ELRA Workshop at MT Summit XI, 2007

[ELDA]

  As you may know, ELRA is  active in the field of evaluation. . In this 
context, ELRA announces a workshop:
 
ELRA Workshop at MT Summit XI, 2007
=============================
 
AUTOMATIC PROCEDURES IN MT EVALUATION
=====================================
 
This workshop, during MT Summit XI, Copenhagen 2007 (Sept. 11), focusses 
on the discussion of automatic evaluation procedures in MT: BLEU / NIST, 
d-score, x-score, edit distance, and other such tools.
 
The questions to be discussed are:
 
·        What do the scores really measure? Are they biased towards 
specific MT technologies? (validity)
·        What kind initial effort do they require (e.g.: pre-translate 
test corpus)? (economy)
·        What kind of implicit assumptions do they make?
·        What kind of resources do they need (e.g.: third party 
grammars)? (economy, feasibility)
·        What kind of diagnostic support can they give? (where to 
improve the system)
·        What kind of evaluation criteria (related to the FEMTI 
framework) do they support (adequacy, fluency, ...)
 
The objective of the workshop is to learn from recent evaluation 
activities, and to  create a better understanding of the strengths and 
limitations of the respective approaches, and to get closer to a common 
methodology for MT output evaluation.
 

Draft programme
 
9.00 Welcome and introduction
 
9.20 The place of automatic evaluation metrics in external quality 
models for
machine translation
Andrei Popescu-Belis, University of Geneva
 
10.00 Evaluating Evaluation --- Lessons from the WMT'07 Shared Task
Philipp Koehn, University of Edinburgh
 
10.30 Coffee break
11.00 Investigating Why BLEU Penalizes Non-Statistical Systems
Eduard Hovy, University of Southern California
 
11.30 Edit distance as an evaluation metric
Christopher Cieri, Linguistic Data Consortium (TBC)
 
12.00 Experience and conclusions from the CESTA evaluation project
Olivier Hamon, ELDA
 
12.30 Lunch
13.30 Automatic Evaluation in MT system production
Gregor Thurmair, Linguatec
 
14.00 Sensitivity of performance-based and proximity-based models for MT 
evaluation
Bogdan Babych, Univ. Leeds
 
14.30 Automatic & human Evaluations of MT in the framework of a speech to
speech communication
Khalid. Choukri, ELDA
 
15.00 Coffee break
15.30 Discussion and conclusions
17.00 Close
 
More information will be found under the MT Summit website:   
http://mtsummitcph.ku.dk
 
Kindest regards,
ELRA evaluation committee
(B. Maegaard, Kh. Choukri, Gr. Thurmair)