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