| # **Please access the latest verison of data that is here https://huggingface.co/datasets/shainar/BEAD ** | |
| email at shaina.raza@torontomu.ca for usage of data | |
| # Please cite us if you use it | |
| ```bibtex | |
| @article{raza2024beads, | |
| title={BEADs: Bias Evaluation Across Domains}, | |
| author={Raza, Shaina and Rahman, Mizanur and Zhang, Michael R}, | |
| journal={arXiv preprint arXiv:2406.04220}, | |
| year={2024} | |
| } | |
| ``` | |
| --- | |
| license: cc-by-nc-4.0 | |
| language: | |
| - en | |
| pretty_name: Navigating News Narratives | |
| task_categories: | |
| - text-classification | |
| - token-classification | |
| --- | |
| # Navigating News Narratives: A Media Bias Analysis Dataset | |
| ## Introduction | |
| The growing prevalence of bias in news media critically influences public perceptions on various imperative subjects, encompassing areas like political standpoints, health, resource allocation, racial perspectives, age, gender biases, and climate change. In light of this, it's crucial for media outlets to uphold the ethics of providing precise information and heightening public cognizance about these pivotal issues. The necessity for a mechanism that could counteract the spread of fallacious or misleading details is thus underscored, aiming to rejuvenate public confidence in the media. | |
| ## Dataset Description | |
| This dataset encompasses multiple dimensions of biases in news media, such as political inclinations, hate speech, toxicity, sexism, ageism, and more, establishing its distinctiveness in the realm of similar datasets. It's noteworthy that the dataset explicitly refrains from including any personally identifiable information (PII). | |
| ### Format | |
| The data structure is tabulated as follows: | |
| - **Text**: The main content. | |
| - **Dimension**: Descriptive category of the text. | |
| - **Biased_Words**: A compilation of words regarded as biased. | |
| - **Aspect**: Specific sub-topic within the main content. | |
| - **Label**: Indicates the presence (True) or absence (False) of bias. The label is ternary - highly biased, slightly biased and neutral | |
| - **Toxicity**: Indicates the presence (True) or absence (False) of bias. | |
| - **Identity_mention**: Mention of any identity based on words match. | |
| ### Annotation Scheme | |
| The labels and annotations in the dataset are generated through a system of Active Learning, cycling through: | |
| - Manual Labeling | |
| - Semi-Supervised Learning | |
| - Human Verification | |
| The scheme comprises: | |
| - **Bias Label**: Specifies the degree of bias (e.g., no bias, mild, or strong). | |
| - **Words/Phrases Level Biases**: Pinpoints specific biased terms or phrases. | |
| - **Subjective Bias (Aspect)**: Highlights biases pertinent to content dimensions. | |
| *Due to the nuances of semantic match algorithms, certain labels such as 'identity' and 'aspect' may appear distinctively different.* | |
| ## Datasets Utilized | |
| For a comprehensive perspective, various news categories like climate crisis summaries, occupational, spiritual/faith, etc., have been curated using RSS. Active learning has been harnessed to classify sentences based on their neutrality or degree of bias and identify biased terms. | |
| Moreover, this compilation incorporates data from several reputable sources: | |
| - **MBIC (media bias)**: [Source](https://zenodo.org/records/4474336) | |
| - **Hyperpartisan news**: [Source](https://huggingface.co/datasets/hyperpartisan_news_detection) | |
| - **Toxic comment classification**: [Kaggle Source](https://kaggle.com/competitions/jigsaw-toxic-comment-classification-challenge) | |
| - **Jigsaw Unintended Bias**: [Kaggle Source](https://kaggle.com/competitions/jigsaw-unintended-bias-in-toxicity-classification) | |
| - **Age Bias**: [Harvard Source](https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/L4OAKN) | |
| - **Multi-dimensional news (Ukraine)**: [Source](https://zenodo.org/records/3885351#.ZF0KoxHMLtV) | |
| - **Social biases**: [Source](https://maartensap.com/social-bias-frames/) | |
| ## Objective | |
| This dataset aims to provide unrestricted, complimentary access, striving to facilitate its adoption among global AI researchers and practitioners. Emphasis is placed on ensuring easy accessibility and usage of the dataset. | |
| https://arxiv.org/abs/2312.00168 | |
| ## Licensing & Citation | |
| If you leverage this dataset in your work, we kindly request you to cite us: | |
| > Navigating News Narratives: A Media Bias Analysis Dataset © 2023 by Shaina Raza, Vector Institute is licensed under [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/). | |
| For any other datasets utilized in this project, please ensure appropriate attribution as detailed in the aforementioned list. | |
| [Hugging Face Source](https://huggingface.co/datasets/newsmediabias/news-bias-full-data/edit/main/README.md) - Accessed on 10/24/2023. | |
| ## Citation | |
| ```bibtex | |
| @misc{Raza2023MediaBiasDataset, | |
| title={Navigating News Narratives: A Media Bias Analysis Dataset}, | |
| author={Shaina Raza}, | |
| year={2023}, | |
| note={Licensed under CC BY-NC 4.0}, | |
| url={https://creativecommons.org/licenses/by-nc/4.0/} | |
| } | |
| ``` | |
| ```bibtex | |
| @article{raza2023navigating, | |
| title={Navigating News Narratives: A Media Bias Analysis Dataset}, | |
| author={Raza, Shaina}, | |
| journal={arXiv preprint arXiv:2312.00168}, | |
| year={2023} | |
| } | |
| ``` |