README.md

metadata

license: odc-by
task_categories:
  - text-generation
dataset_info:
  - config_name: all
    features:
      - name: section
        dtype: string
      - name: filename
        dtype: string
      - name: text
        dtype: string
      - name: lang
        dtype: string
    splits:
      - name: test
        num_bytes: 193159922
        num_examples: 4334
      - name: validation
        num_bytes: 212095720
        num_examples: 4328
      - name: train
        num_bytes: 7445907253
        num_examples: 159185
    download_size: 2899594863
    dataset_size: 7851162895
  - config_name: all-filtered
    features:
      - name: filename
        dtype: string
      - name: text
        dtype: string
      - name: flesch_reading_ease
        dtype: float64
    splits:
      - name: test
        num_bytes: 157823888.30671677
        num_examples: 3549
      - name: validation
        num_bytes: 173010266.388928
        num_examples: 3541
      - name: train
        num_bytes: 6018286403.457957
        num_examples: 129653
    download_size: 2428198753
    dataset_size: 6349120558.153602
  - config_name: all-filtered-deduped
    features:
      - name: filename
        dtype: string
      - name: text
        dtype: string
      - name: flesch_reading_ease
        dtype: float64
    splits:
      - name: train
        num_bytes: 5593504495.685427
        num_examples: 121120
      - name: validation
        num_bytes: 168820995.24710205
        num_examples: 3403
      - name: test
        num_bytes: 141318723.08888263
        num_examples: 3417
    download_size: 1469473020
    dataset_size: 5903644214.021411
  - config_name: doc
    features:
      - name: section
        dtype: string
      - name: filename
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 3373404810.354176
        num_examples: 66804
      - name: validation
        num_bytes: 88773810.7987941
        num_examples: 1758
      - name: test
        num_bytes: 88824307.84703004
        num_examples: 1759
    download_size: 1362475698
    dataset_size: 3551002929.0000005
  - config_name: docx
    features:
      - name: section
        dtype: string
      - name: filename
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 4605598.853503184
        num_examples: 141
      - name: validation
        num_bytes: 261310.57324840763
        num_examples: 8
      - name: test
        num_bytes: 261310.57324840763
        num_examples: 8
    download_size: 1788590
    dataset_size: 5128220
  - config_name: embeddings-jina-base
    features:
      - name: filename
        dtype: string
      - name: text
        dtype: string
      - name: flesch_reading_ease
        dtype: float64
      - name: embedding
        sequence: float64
    splits:
      - name: train
        num_bytes: 6327123534
        num_examples: 125834
      - name: validation
        num_bytes: 189970363
        num_examples: 3537
      - name: test
        num_bytes: 161101531
        num_examples: 3544
    download_size: 2921652603
    dataset_size: 6678195428
  - config_name: logs
    features:
      - name: section
        dtype: string
      - name: filename
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 2350324475.916881
        num_examples: 9223
      - name: validation
        num_bytes: 61924411.541559376
        num_examples: 243
      - name: test
        num_bytes: 61924411.541559376
        num_examples: 243
    download_size: 718096901
    dataset_size: 2474173298.9999995
  - config_name: ppt
    features:
      - name: section
        dtype: string
      - name: filename
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: validation
        num_bytes: 11813294
        num_examples: 1230
      - name: train
        num_bytes: 426593595
        num_examples: 43706
      - name: test
        num_bytes: 12242562
        num_examples: 1232
    download_size: 232304159
    dataset_size: 450649451
  - config_name: pptx
    features:
      - name: section
        dtype: string
      - name: filename
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 9517778
        num_examples: 963
      - name: validation
        num_bytes: 513930
        num_examples: 53
      - name: test
        num_bytes: 436852
        num_examples: 54
    download_size: 5314310
    dataset_size: 10468560
  - config_name: rtf
    features:
      - name: section
        dtype: string
      - name: filename
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 61558658.13180516
        num_examples: 942
      - name: validation
        num_bytes: 3398142.4871060173
        num_examples: 52
      - name: test
        num_bytes: 3463491.3810888254
        num_examples: 53
    download_size: 22547280
    dataset_size: 68420292
  - config_name: txt
    features:
      - name: section
        dtype: string
      - name: filename
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 1358006724.1432111
        num_examples: 41393
      - name: validation
        num_bytes: 35727522.10740843
        num_examples: 1089
      - name: test
        num_bytes: 35760329.749380335
        num_examples: 1090
    download_size: 608912009
    dataset_size: 1429494576
configs:
  - config_name: all
    data_files:
      - split: test
        path: all/test-*
      - split: validation
        path: all/validation-*
      - split: train
        path: all/train-*
  - config_name: all-filtered
    data_files:
      - split: test
        path: all-filtered/test-*
      - split: validation
        path: all-filtered/validation-*
      - split: train
        path: all-filtered/train-*
  - config_name: all-filtered-deduped
    data_files:
      - split: train
        path: all-filtered-deduped/train-*
      - split: validation
        path: all-filtered-deduped/validation-*
      - split: test
        path: all-filtered-deduped/test-*
  - config_name: doc
    data_files:
      - split: train
        path: doc/train-*
      - split: validation
        path: doc/validation-*
      - split: test
        path: doc/test-*
  - config_name: docx
    data_files:
      - split: train
        path: docx/train-*
      - split: validation
        path: docx/validation-*
      - split: test
        path: docx/test-*
  - config_name: embeddings-jina-base
    data_files:
      - split: train
        path: embeddings-jina-base/train-*
      - split: validation
        path: embeddings-jina-base/validation-*
      - split: test
        path: embeddings-jina-base/test-*
  - config_name: logs
    data_files:
      - split: train
        path: logs/train-*
      - split: validation
        path: logs/validation-*
      - split: test
        path: logs/test-*
  - config_name: ppt
    data_files:
      - split: validation
        path: ppt/validation-*
      - split: train
        path: ppt/train-*
      - split: test
        path: ppt/test-*
  - config_name: pptx
    data_files:
      - split: train
        path: pptx/train-*
      - split: validation
        path: pptx/validation-*
      - split: test
        path: pptx/test-*
  - config_name: rtf
    data_files:
      - split: train
        path: rtf/train-*
      - split: validation
        path: rtf/validation-*
      - split: test
        path: rtf/test-*
  - config_name: txt
    data_files:
      - split: train
        path: txt/train-*
      - split: validation
        path: txt/validation-*
      - split: test
        path: txt/test-*

govdocs1 Dataset: By File Extension

Markdown-parsed versions of documents in govdocs1 with light filtering.

Usage

Load specific file formats (e.g., .doc files) parsed to markdown with pandoc:

from datasets import load_dataset

# Replace "doc" with desired config name
dataset = load_dataset("BEE-spoke-data/govdocs1-by-extension", "doc")

Configurations

This dataset includes multiple configurations, each corresponding to a different file extension:

• doc
• docx
• logs
• ppt
• pptx
• rtf
• txt

Each configuration contains train, validation, and test splits.

Dataset Details

• Download Size: Varies by configuration
• Dataset Size: Varies by configuration
• Splits: Train, Validation, Test
• Features: Section, Filename, Text

counts

Here's a summary of the number of examples for each configuration in the dataset (thus far):

1. DOC Configuration

   • Train Examples: 38,094
   • Validation Examples: 1,002
   • Test Examples: 1,003

2. DOCX Configuration

   • Train Examples: 141
   • Validation Examples: 8
   • Test Examples: 8

3. Logs Configuration

   • Train Examples: 9,223
   • Validation Examples: 243
   • Test Examples: 243

4. PPT Configuration

   • Train Examples: 13,865
   • Validation Examples: 365
   • Test Examples: 365

5. PPTX Configuration

   • Train Examples: 963
   • Validation Examples: 53
   • Test Examples: 54

6. RTF Configuration

   • Train Examples: 942
   • Validation Examples: 52
   • Test Examples: 53

7. TXT Configuration

   • Train Examples: 41,393
   • Validation Examples: 1,089
   • Test Examples: 1,090

Citation

@inproceedings{garfinkel2009bringing,
  title={Bringing Science to Digital Forensics with Standardized Forensic Corpora},
  author={Garfinkel, Simson and others},
  booktitle={Digital Forensic Research Workshop (DFRWS) 2009},
  year={2009},
  address={Montreal, Canada},
  url={https://digitalcorpora.org/corpora/file-corpora/files/}
}

For more detailed information on each configuration, refer to the dataset documentation.