metadata

language:
  - en
license:
  - cc0-1.0
size_categories:
  - 10K<n<100K
tags:
  - government
  - declassified
  - psychic-research
  - remote-viewing
  - historical-documents

Dataset Card for STARGATE Project Documents

Dataset Description

Dataset Summary

The STARGATE Project Documents dataset contains declassified documents from the U.S. government's psychic phenomena research program that ran from the 1970s to the 1990s. This collection includes text extracted from PDF files, with metadata and categorization to facilitate research and analysis.

Languages

The documents are primarily in English (en).

Dataset Structure

Data Instances

Each document in the dataset includes:

The full text content extracted from PDFs
Metadata including document ID, classification, dates, and named entities
Category assignment based on document content

Data Fields

filename: Original filename
document_id: CIA document identifier
classification: Security classification level
category: Document category/type
word_count: Number of words in the document
page_count: Estimated number of pages
dates: Dates mentioned in the document
project_names: Names of projects mentioned
persons: Person names identified in the document
organizations: Organization names identified
locations: Location names identified

Data Splits

The dataset does not have formal splits, but documents are organized by category.

Dataset Creation

Source Data

The source data consists of declassified PDF documents from the CIA's CREST database, specifically those related to the STARGATE project and related psychic research programs.

Annotations

The dataset includes automatically extracted metadata and categorization based on document content analysis.

Considerations for Using the Data

Social Impact of Dataset

This dataset provides transparency into historical government research programs and facilitates academic study of these programs.

Discussion of Biases

The documents reflect the perspectives and biases of their time and the institutions that created them.

Other Known Limitations

Some documents may have incomplete text extraction due to poor scan quality or redactions in the original documents.

Additional Information

Dataset Curators

This dataset was curated by GotThatData.

Licensing Information

These documents are in the public domain as they have been declassified by the U.S. government.

Citation Information

@dataset{stargate_documents,
  author = {GotThatData},
  title = {STARGATE Project Documents},
  year = {2025},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/GotThatData/STARGATE}
}