A set of documents from Reuters' 1986 newswire which have been classified. This dataset is appropriate for testing natural language processing and information retrieval...

This is a publicly available, tokenized version of the Reuters RCV1 corpus by David D Lewis et al. The creator requests attribution.

The ClueWeb09 dataset was created to support research on information retrieval and related human language technologies. It consists of about 1 billion web pages in ten languages...