gtfintechlab/ipo-text

---
license: cc-by-4.0
task_categories:
- text-classification
- sentence-similarity
- document-question-answering
size_categories:
- 100K<n<1M
language:
- en
tags:
- finance
- nlp
- text
configs:
- config_name: default
data_files:
- split: train
path:
- "metadata.jsonl"
dataset_info:
features:
- name: company
dtype: string
- name: company_tickers
sequence: string
- name: cik
dtype: string
- name: sic
dtype: string
- name: office
dtype: string
- name: industry
dtype: string
- name: exchanges
sequence: string
- name: filing_type
dtype: string
- name: filing_date
dtype: string
- name: filing_accession_number
dtype: string
- name: filing_processed
dtype: bool
- name: parsed_toc_keys
dtype: string
- name: parsed_prospectus_summary_valid
dtype: bool
- name: parsed_prospectus_summary
dtype: string
- name: parsed_prospectus_summary_word_count
dtype: int64
- name: parsed_risk_factors_valid
dtype: bool
- name: parsed_risk_factors
dtype: string
- name: parsed_risk_factors_word_count
dtype: int64
- name: parsed_legal_matters_valid
dtype: bool
- name: parsed_legal_matters
dtype: string
- name: parsed_legal_matters_word_count
dtype: int64
- name: filing_url
dtype: string
- name: filing_images
sequence: string
pretty_name: SEC IPO Filings Datase
---

# SEC IPO Filings Dataset
A large-scale, comprehensive dataset of **100,000+ filings** (S-1 and F-1 filings) filed with the [SEC EDGAR](https://www.sec.gov/edgar) system, spanning **1994–2026** and over **20,000** unique registrants.

Every filing has been downloaded and then parsed using the [IPO-Mine Python Package](https://pypi.org/project/ipo-mine/). We have extracted three common sections found in these documents (Prospectus Summary, Risk Factors, Legal Matters), and then used an LLM classifier to group them into three categories. For this dataset, we have only included the parsed sections labeled as "Safe to Use" as valid.

## Software Pipeline
<p align="center">
<img src="assets/software-pipeline.png" alt="Dataset Construction Pipeline" width="600">
</p>

## Citation

The dataset is a product of the following research. If you use this dataset in your work, please cite:

```bibtex
@misc{galarnyk2026ipomine,
title = {IPO-Mine: A Toolkit and Dataset for Section-Structured Analysis of Long, Multimodal IPO Documents},
author = {Galarnyk, Michael and Lohani, Siddharth and Nandi, Sagnik and Patel, Aman and Kannan, Vidhyakshaya and Banerjee, Prasun and Routu, Rutwik and Ye, Liqin and Hiray, Arnav and Somani, Siddhartha and Chava, Sudheer},
year = {2026},
url = {[https://huggingface.co/datasets/gtfintechlab/ipo-images](https://huggingface.co/datasets/gtfintechlab/ipo-images)},
note = {Preprint/Working Paper}
}