Projekter pr. år
Beskrivelse
Dataset Summary
HiFi-KPI is a large-scale dataset designed for financial numerical key performance indicator (KPI) extraction from earnings filings. It is derived from iXBRL filings mandated by the SEC, featuring hierarchical labels structured from the XBRL taxonomy. The dataset consists of ∼1.8M paragraphs and ∼5M entities, each linked to labels in the iXBRL calculation and presentation taxonomies.
Languages
The dataset is in English, extracted from SEC 10-K and 10-Q filings.
| Dato for tilgængelighed | 21 feb. 2025 |
|---|---|
| Forlag | Hugging Face |
Projekter
- 1 Igangværende
-
Guarantees of Factuality in LLM-based Extraction of Financial KPIs from Earnings Transcripts
Bjerva, J. (PI (principal investigator)), Jensen, R. T. A. (Projektdeltager), Rizzi, G. (Supervisor) & Larsen, T. (Supervisor)
01/09/2024 → 31/08/2027
Projekter: Projekt › Forskning
Publikation
- 1 Preprint
-
HiFi-KPI: A Dataset for Hierarchical KPI Extraction from Earnings Filings
Aavang, R. T., Rizzi, G., Bøggild, R., Iolov, A., Zhang, M. & Bjerva, J., feb. 2025, (Afsendt).Publikation: Working paper/Preprint › Preprint