News Articles Labeled for Political Bias Testing

Cs 230 News V2 Test Liberal contains news data for analyzing political leanings. The dataset, created by myradeng, was uploaded to Hugging Face in May 2022. Platform tags indicate it contains text and tabular data, with an estimated size between 10,000 and 100,000 rows.

Use Cases

• Train a classifier to predict the 'liberal' label from article text and metadata.
• Analyze linguistic features in the text column to identify partisan language patterns.
• Benchmark fairness metrics in text classification models using politically labeled data.
• Perform exploratory data analysis on the tabular features to understand dataset composition.

Strengths

• Dataset size estimated in the 10K-100K row range based on platform tags.
• Includes both text modality and tabular structure for multifaceted analysis.

Limitations

• No column names or sample data are provided, hindering immediate understanding of features.
• Data is from 2022 and may not reflect current media or political discourse.
• Potential label bias as the title suggests a focus on 'Liberal' classification.

Provenance

Source

Hugging Face user myradeng.

Collection Method

Collection method unknown; likely assembled for a machine learning course or project (based on 'Cs 230' in title).

Time Range

null

Freshness

Last updated in May 2022; static dataset with no stated update frequency.

Geography

Platform tags include 'Regionus', suggesting a United States focus.