metadata

language:
  - en
dataset_info:
  - config_name: default
    features:
      - name: text
        dtype: string
      - name: label
        dtype:
          class_label:
            names:
              '0': non-cybersec
              '1': cybersec
      - name: source
        dtype: string
    splits:
      - name: train
        num_bytes: 9751954642
        num_examples: 9273685
      - name: validation
        num_bytes: 91058728
        num_examples: 459115
    download_size: 4960015752
    dataset_size: 9843013370
  - config_name: length_filtered
    features:
      - name: text
        dtype: string
      - name: label
        dtype:
          class_label:
            names:
              '0': non-cybersec
              '1': cybersec
      - name: source
        dtype: string
      - name: tokens_length
        dtype: int64
      - name: text_length
        dtype: int64
    splits:
      - name: train
        num_bytes: 4928000253.748813
        num_examples: 4616079
      - name: validation
        num_bytes: 527264.9276978535
        num_examples: 2460
    download_size: 3300180199
    dataset_size: 4928527518.676511
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
  - config_name: length_filtered
    data_files:
      - split: train
        path: length_filtered/train-*
      - split: validation
        path: length_filtered/validation-*

Cybersecurity Topic Classification (CTC) Dataset

Note: This is an unofficial upload of the Cybersecurity Topic Classification (CTC) dataset. The original dataset and accompanying paper were developed by Elijah Pelofske, Lorie M. Liebrock, and Vincent Urias.

This dataset comprises training and validation data for the Cybersecurity Topic Classification (CTC) tool, as introduced in the paper "A Robust Cybersecurity Topic Classification Tool" by Elijah Pelofske, Lorie M. Liebrock, and Vincent Urias. The dataset facilitates the detection of cybersecurity-related discussions in natural English text.

Dataset Overview

The dataset includes unprocessed English text samples labeled as either cybersecurity-related or non-cybersecurity-related. These samples were sourced from Reddit, StackExchange sites, and arXiv documents, utilizing user-defined tags and metadata for labeling. Please note that the data has not undergone manual labeling, post-processing, filtering, or censorship; therefore, some labels may be incorrect.

Accessing the Data

You can load this dataset directly using the Hugging Face datasets library:

from datasets import load_dataset

dataset = load_dataset("naufalso/cybersec-topic-classification-dataset")

Usage

This dataset is intended for research purposes, particularly in training and evaluating machine learning models for cybersecurity topic classification. Users should be aware of the potential for incorrect labels and the unprocessed nature of the text data.

Citation

If you utilize this dataset in your research, please cite the original paper:

@article{pelofske2021robust,
  title={A Robust Cybersecurity Topic Classification Tool},
  author={Pelofske, Elijah and Liebrock, Lorie M. and Urias, Vincent},
  journal={arXiv preprint arXiv:2109.02473},
  year={2021}
}

Acknowledgments

We extend our gratitude to the authors for providing this dataset and to the platforms Reddit, StackExchange, and arXiv for making their data publicly accessible.

For additional information and resources, please refer to the original GitHub repository.