Skip to page content
SAVRN
Search Plan a project

Dataset · Text classification

TRACE-OU-Bench

by TRACE-OU TRACE-OU/TRACE-OU-Bench

TRACE-OU-Bench is a sequential benchmark for ontology-update decisions under evolving ontology states. It contains 620 human-reviewed ontology-update proposals generated from public U.S. SEC 10-Q filings covering 45 U.S.

Rows870
Configurations2
Size15.1 MB
License—
AccessPublicly accessible
Monthly Downloads54

Dataset Card

TRACE-OU-Bench is a sequential benchmark for ontology-update decisions under evolving ontology states. It contains 620 human-reviewed ontology-update proposals generated from public U.S. SEC 10-Q filings covering 45 U.S. companies, five sectors, and 2016–2026. The public release does not redistribute SEC filing text. Instead, each record retains SEC provenance, deterministic chunk identifiers, offsets, and SHA-256 hashes. reconstruction/traceoureconstructv11.py downloads the original filing from SEC EDGAR and deterministically reconstructs the source chunk. - Train (2016–2023): 350 proposals generated against O0; reviewed Train updates build O1. - Test (2025–2026): 150 proposals generated…

Excerpt from the card by TRACE-OU.

Structure

default 620 rows

SplitRowsSize
train3502.5 MB
calibration1203.0 MB
test1503.8 MB
candidate_idstringsplitstringsnapshot_idstringsnapshot_fingerprintstringchunk_idstringdocument_idstringtickerstringcompany_namestringindustrystringfiling_datestringsectionstringsource_urlstringaccession_numberstringsection_quality_flagstringchunk_indexint64chunk_word_startint64chunk_word_end_exclusiveint64word_countint64chunk_sha256stringevidence_alignment_statusstringevidence_source_char_startfloat64evidence_source_char_end_exclusivefloat64evidence_source_word_startfloat64evidence_source_word_end_exclusivefloat64evidence_source_sha256stringevidence_quote_sha256stringactionstringrelationstringsubject_typestringobject_typestringtarget_relation_idstringsecondary_relation_idstringtemporal_expressionstringrationalestringraw_confidencefloat64hard_validation_passint64generator_modelstringgenerator_versionstringcreated_atstringreport_datestring

prefilter_audit 250 rows

SplitRowsSize
train100622.9 KB
calibration751.8 MB
test751.8 MB
splitstringsnapshot_idstringsnapshot_fingerprintstringchunk_idstringdocument_idstringtickerstringcompany_namestringindustrystringfiling_datestringsectionstringsource_urlstringaccession_numberstringchunk_indexint64chunk_word_startint64chunk_word_end_exclusiveint64word_countint64chunk_sha256stringhybrid_scorefloat64prefilter_selectedint64audit_stratumstringsnapshot_relation_countint64snapshot_relation_inventory_textstringontology_context_jsonstringaudit_weightfloat64audit_contains_updateint64audit_candidate_coveragestringcandidate_summary_jsonstring

Details

Repository
TRACE-OU/TRACE-OU-Bench
Publisher
TRACE-OU
Task category
Text classification
Tags
knowledge-graph, ontology, ontology-evolution
Size category
n<1K
Languages
en
Revision
1ae1dbaa2aab85205c80e12904e27210157c8682
Last updated
2026-10-06

Files

38 files, 15.1 MB in total.

Data22 files · 15.0 MB
Documentation11 files · 55.2 KB
Other4 files · 56.9 KB
Repository1 file · 2.5 KB
Every file
FileTypeSizeSHA-256
SHA256SUMS.txtData3.6 KB—
audit/calibration_prefilter_audit.csvData2.0 MB—
audit/test_prefilter_audit.csvData2.0 MB—
audit/train_prefilter_audit.csvData672.5 KB—
data/calibration.csvData3.3 MB—
data/test.csvData4.1 MB—
data/train.csvData2.6 MB—
ontology/O0_relation_inventory.txtData1.5 KB—
ontology/O0_relations.jsonData4.9 KB—
ontology/O0_seed_ontology.jsonData14.4 KB—
ontology/O1_relation_inventory.txtData6.9 KB—
ontology/O1_relations.jsonData19.7 KB—
ontology/O2_relation_inventory.txtData6.9 KB—
ontology/O2_relations.jsonData19.9 KB—
ontology/snapshot_manifest.jsonData1.2 KB—
publishing/viewer_configuration_validation.jsonData508 B—
reconstruction/full_reconstruction_validation.jsonData8.0 KB—
reconstruction/parser_compatibility_validation.jsonData8.1 KB—
reconstruction/reconstruction_fix_validation.jsonData1.4 KB—
reconstruction/release_chunk_manifest.csvData243.1 KB—
reconstruction/requirements.txtData76 B—
reliability/blinded_repeat_mapping.csvData6.8 KB—
DATA_DICTIONARY.mdDocumentation8.5 KB—
README.mdDocumentation14.3 KB—
SOURCE_RECONSTRUCTION_NOTICE.mdDocumentation389 B—
SOURCE_TEXT_MIGRATION_REPORT.mdDocumentation1.5 KB—
VALIDATION.mdDocumentation5.0 KB—
publishing/HUGGINGFACE_UPDATE_ZH.mdDocumentation5.2 KB—
reconstruction/CANONICAL_SPEC.mdDocumentation3.6 KB—
reconstruction/FULL_VALIDATION_ZH.mdDocumentation3.9 KB—
reconstruction/PARSER_VERSION_FIX_ZH.mdDocumentation3.3 KB—
reconstruction/README_RECONSTRUCTION.mdDocumentation6.6 KB—
reconstruction/RECONSTRUCTION_FIX_ZH.mdDocumentation2.8 KB—
publishing/upload_hf_release.pyOther5.2 KB—
reconstruction/test_reconstruction.pyOther11.5 KB—
reconstruction/trace_ou_reconstruct_v11.pyOther27.8 KB—
reconstruction/validate_reconstructed_release.pyOther12.4 KB—
.gitattributesRepository2.5 KB—

License and Download

License
Not stated by the source
Access
No access gate
Download from TRACE-OU

Released by TRACE-OU through its official repository on Hugging Face.