license: apache-2.0
task_categories:
- text-classification
language:
- en
tags:
- finance
size_categories:
- 1K<n<10K
Dataset Card for Financial Fraud Labeled Dataset
Dataset Details
This dataset collects financial filings from various companies submitted to the U.S. Securities and Exchange Commission (SEC). The dataset consists of 85 companies involved in fraudulent cases and an equal number of companies not involved in fraudulent activities. The Fillings column includes information such as the company's MD&A, and financial statement over the years the company stated on the SEC website.
This dataset was used for research in detecting financial fraud using multiple LLMs and traditional machine-learning models.
- Curated by: Amit Kedia
- Language(s) (NLP): English
- License: Apache 2.0
Dataset Sources
- Repository: GitHub
- Thesis: Financial Fraud Detection using LLMs
Direct Use
Code to Directly use the dataset:
from datasets import load_dataset
dataset = load_dataset("amitkedia/Financial-Fraud-Dataset")
Out-of-Scope Use
There are some limitations of the dataset:
- This dataset is designed for acedemic research
- The text needs to be cleaned for further process
- The dataset does not cover all the fradulent cases and are limited to Securities and Exchange Commision of USA (SEC) that means the fradulent and non fradulent cases are the companies of USA
Dataset Structure
For the structure of the dataset look into the dataset viewer.
Dataset Creation
Check out the Thesis
Curation Rationale
To help the financial industry develop the best model to detect fraudulent activities which can save billions of dollars for government and banks
Data Collection and Processing
Please Refer to the Thesis