IARPA BETTER (Better Extraction from Text Towards Enhanced Retrieval) information extraction and information retrieval datasets.
Cross-language information extraction and retrieval datasets developed for the evaluation of the IARPA BETTER program. The documents come from CommonCrawl. The IE annotations in three schemas are by MITRE and ARLIS. The IR queries and relevance judgments were done at NIST, and NIST was asked by IARPA to distribute the data in its final form. The tasks are all cross-language from English into one of Arabic, Farsi, Russian, Chinese, and Korean
Complete Metadata
| @type | dcat:Dataset |
|---|---|
| accessLevel | public |
| accrualPeriodicity | irregular |
| bureauCode |
[
"006:55"
]
|
| contactPoint |
{
"fn": "Ian Soboroff",
"hasEmail": "mailto:ian.soboroff@nist.gov"
}
|
| description | Cross-language information extraction and retrieval datasets developed for the evaluation of the IARPA BETTER program. The documents come from CommonCrawl. The IE annotations in three schemas are by MITRE and ARLIS. The IR queries and relevance judgments were done at NIST, and NIST was asked by IARPA to distribute the data in its final form. The tasks are all cross-language from English into one of Arabic, Farsi, Russian, Chinese, and Korean |
| distribution |
[
{
"title": "The BETTER datasets",
"mediaType": "application/octet-stream",
"downloadURL": "https://ir.nist.gov/better/"
}
]
|
| identifier | ark:/88434/mds2-2946 |
| issued | 2024-04-22 |
| keyword |
[
"information extraction; information retrieval; cross-language information retrieval"
]
|
| landingPage | https://ir.nist.gov/better/ |
| language |
[
"en"
]
|
| license | https://www.nist.gov/open/license |
| modified | 2023-02-24 00:00:00 |
| programCode |
[
"006:045"
]
|
| publisher |
{
"name": "National Institute of Standards and Technology",
"@type": "org:Organization"
}
|
| theme |
[
"Information Technology:Data and informatics"
]
|
| title | IARPA BETTER (Better Extraction from Text Towards Enhanced Retrieval) information extraction and information retrieval datasets. |
