Once imported your files are automatically classified and the URL and access date added as attributes. PDFs and Facebook and Twitter datasets are assigned to the 'Reference' classification. YouTube videos and datasets are assigned to the 'YouTube' classification.