Data
|
Description
|
Size
|
Due Date
|
Delivered |
Notes
|
HARD 2004 Evaluation Corpus
|
English only newswire 5 sources, January-December 2003
|
|
mid April, 2004
|
May 21, 2004
|
LDC2004E30
Contact ldc@ldc.upenn.edu
|
Training topics
|
20 new or revised topic descriptions, with metadata
|
20 topic description text files
|
May 20, 2004
|
June 14, 2004
|
LDC2004E32 Contact ldc@ldc.upenn.edu
|
Evaluation topics
|
50 new topic descriptions with metadata
|
50 topic description files
|
June 25, 2004
|
June 25, 2004
|
|
Clarification forms
|
LDC returns HTML clarification forms, received July 16,2004, to research sites.
|
up to 3 forms per topic, per site
|
July 29, 2004 |
July 23, 2004 |
Emailed directly to sites |
Baseline runs |
Sites deliver baseline runs (no metadata) to NIST
|
n/a |
July 9, 2004
|
n/a |
received |
Final runs
|
Sites deliver revised/final runs (metadata) to NIST
|
n/a
|
August 5, 2004
|
n/a |
received |
Evaluation relevance judgements
|
LDC assesses pooled lists of documents for each topic
|
85 pooled documents per site per topic
|
October 11, 2004
|
delivered to NIST
|
LDC2004E42 Contact ldc@ldc.upenn.edu
|
TREC meeting |
Principal Investigator's meeting, Gaithersburg, MD |
participating sites
|
November 16-19, 2004
|
n/a
|
|