2,225 labeled news articles from the BBC website spanning the years 2004 to 2005. The data is organized into five thematic categories: business, entertainment, politics, sport, and tech.
Use Cases
- Train a text classification model to map article content to one of the five category labels
- Benchmark document clustering techniques using the provided category labels as ground truth
- Conduct temporal analysis of news topics within the 2004-2005 timeframe using the article text
Strengths
- 2,225 news articles published between 2004 and 2005
- Five distinct category labels: business, entertainment, politics, sport, and tech
- Sourced from the ICML 2006 paper on diagonal dominance in kernel document clustering