A collection of just over 4,000 unique English-language news articles scraped from Arise TV, a popular Nigerian television news source. It is intended for text classification tasks, such as categorizing articles into different topics.
Use Cases
- Classify news articles into categories using the 'Title' text field for topic modeling.
- Analyze the distribution of news topics over time, leveraging the dataset's temporal metadata.
- Train a text classifier on the 'Title' field to automatically categorize new Nigerian news content.
Strengths
- Contains just over 4,000 unique news articles.
- Focuses on a specific geographic source (Nigerian Arise TV).
- Designed for a specific task (news article classification).
Limitations
- Small sample size of just over 4,000 articles, which may limit model generalizability.
- Limited to a single news source, potentially introducing media bias.
- Only the 'Title' column is confirmed; other potential features like full article text are unknown.
Provenance
- Source
- Arise TV, a popular Nigerian news television channel.
- Collection Method
- Scraped from the source website.
- Time Range
- null
- Freshness
- null
- Geography
- Nigeria