Financial News Earnings Information and Evidence Extraction Dataset

#information extraction #earnings attribute extraction #source evidence alignment #financial news analysis #corporate earnings monitoring #financial information verification

5000
0.5GB
JSON
CATL
MOBIUSI INC

Updated:2026-11-03

AI Analysis & Value Prop

This dataset pairs original business and financial news articles with structured JSON annotations covering company names, ticker symbols, reporting periods, financial metrics, and year-over-year or quarter-over-quarter changes. Each extracted item is linked to an exact supporting passage with character offsets and a JSON-style field path, making the source of each result easy to verify. It supports structured-output supervised fine-tuning, financial information extraction, evidence grounding, and related model evaluation.

Dataset Insights

Sample Examples

{}

Technical Specifications

Field	Type	Description
file_name	string	File name
news_text	string	The original, unmodified business or financial news article used as the source for information extraction.
company_name	string	The company whose earnings or financial results are discussed in the article.
ticker_symbol	string	The company's stock ticker symbol as stated or unambiguously identified in the article. Omit this field when unavailable.
evidence_spans	array	Exact source-article spans supporting extracted fields, located with zero-based character offsets and an exclusive end offset, and associated with JSON-style field paths.
reporting_period	object	The fiscal or calendar period to which the reported financial results apply.
financial_metrics	array	Financial results and related changes reported for the company and reporting period, with one array item for each distinct metric.

Compliance Statement

Authorization Type	Proprietary - Commercial AI Training License (No Redistribution)
Commercial Use	Requires exclusive subscription or authorization contract (monthly or per-invocation charging)
Privacy and Anonymization	No PII, no real company names, simulated scenarios follow industry standards
Compliance System	Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs

Frequently Asked Questions

What does the dataset contain?: It includes original business and financial news articles with corresponding JSON annotations for company names, ticker symbols, reporting periods, financial metrics, and changes. Extracted fields are linked to the source passages that support them.

How are evidence spans located?: Each evidence span is an exact passage from the source article. It is located with zero-based character offsets and an exclusive end offset, and associated with a JSON-style field path so users can verify the extraction against the original text.

Which training and evaluation tasks does it support?: The dataset supports extracting earnings information from business and financial news, recognizing financial metrics, generating structured JSON, and evaluating whether a model grounds its extracted results in the correct source evidence.

What are the dataset size and licensing terms?: The supplied information does not specify the sample count or license terms. Before acquisition, review the product listing for dataset size, delivery details, and permitted uses, and confirm that the terms fit your research or commercial needs.

Can't find the data you need?

Post a request and let data providers reach out to you.

Get this Dataset

Verified for Enterprise Use

Cite this Work

@dataset{Mobiusi2026,
  title={Financial News Earnings Information and Evidence Extraction Dataset},
  author={MOBIUSI INC},
  year={2026},
  url={https://www.mobiusi.com/datasets/cf8b95cd8884076a0c5c4fb92076b428},
  urldate={2026-10-20},
  keywords={financial news dataset,earnings extraction,financial metric recognition,news evidence linking,structured output training,company earnings analysis,financial text data,information extraction dataset},
  version={1.0}
}

Using this in research? Please cite us.

placeholder

Scene

Domain

Modality

Task