NIST 2002 Open Machine Translation (OpenMT) Evaluation
|Item Name:||NIST 2002 Open Machine Translation (OpenMT) Evaluation|
|Author(s):||NIST Multimodal Information Group|
|LDC Catalog No.:||LDC2010T10|
|Release Date:||May 14, 2010|
|Language(s):||English, Mandarin Chinese, Standard Arabic, Arabic|
|Language ID(s):||eng, cmn, arb, ara|
LDC User Agreement for Non-Members
|Online Documentation:||LDC2010T10 Documents|
|Licensing Instructions:||Subscription & Standard Members, and Non-Members|
|Citation:||NIST Multimodal Information Group. NIST 2002 Open Machine Translation (OpenMT) Evaluation LDC2010T10. Web Download. Philadelphia: Linguistic Data Consortium, 2010.|
NIST 2002 Open Machine Translation (OpenMT) Evaluation is a package containing source data, reference translations, and scoring software used in the NIST 2002 OpenMT evaluation. It is designed to help evaluate the effectiveness of machine translation systems. The package was compiled and scoring software was developed by researchers at NIST, making use of newswire source data and reference translations collected and developed by LDC.
The objective of the NIST OpenMT evaluation series is to support research in, and help advance the state of the art of, machine translation (MT) technologies -- technologies that translate text between human languages. Input may include all forms of text. The goal is for the output to be an adequate and fluent translation of the original.
The MT evaluation series started in 2001 as part of the DARPA TIDES (Translingual Information Detection, Extraction) program. Beginning with the 2006 evaluation, the evaluations have been driven and coordinated by NIST as NIST OpenMT. These evaluations provide an important contribution to the direction of research efforts and the calibration of technical capabilities in MT. The OpenMT evaluations are intended to be of interest to all researchers working on the general problem of automatic translation between human languages. To this end, they are designed to be simple, to focus on core technology issues, and to be fully supported. The 2002 task was to evaluate translation from Chinese to English and from Arabic to English.
Additional information about these evaluations may be found at the NIST Open Machine Translation (OpenMT) Evaluation web site.
This evaluation kit includes a single perl script (mteval-v09.pl) that may be used to produce a translation quality score for one (or more) MT systems. The script works by comparing the system output translation with a set of (expert) reference translations of the same source text. Comparison is based on finding sequences of words in the reference translations that match word sequences in the system output translation. More information on the evaluation algorithm may be obtained from the paper detailing the algorithm: BLEU: a Method for Automatic Evaluation of Machine Translation (Papineni et al, 2002).
The Chinese-language source text included in this corpus is a reorganization of data that was initially released to the public as Multiple-Translation Chinese (MTC) Part 2 (LDC2003T17). The Chinese-language reference translations are a reorganized subset of data from the same MTC corpus. The Arabic-language data (source text and reference translations) is a reorganized subset of data that was initially released to the public as Multiple-Translation Arabic (MTA) Part 1 (LDC2003T18). All source data for this corpus is newswire text. Chinese source text was drawn in March and April 2002 from Xinhua News Agency and in March 2002 from Zaobao News Service (sources indicated in docids). Arabic source text was drawn from the Xinhua News Agency's Arabic newswire feed (October 2001, in the docid range: artb_500 - artb_565) and Agence France-Presse (Feb. 1998 - Oct. 1999, in the docid range: artb_001 - artb_069). Arabic Agence France-Presse source text was also released as part of Arabic Newswire Part 1 (LDC2001T55). For details on the methodology of the source data collection and production of reference translations, see the documentation for the above-mentioned corpora.
For each language, the test set consists of two files, a source and a reference file. Each reference file contains four independent translations of the data set. The evaluation year, source language, test set (which, by default, is "evalset"), version of the data, and source vs. reference file (with the latter being indicated by "-ref") are reflected in the file name.
DARPA TIDES MT and NIST OpenMT evaluations used SGML-formatted test data until 2008 and XML-formatted test data thereafter. The files in this package are provided in both formats.
No updates have been issued at this time.