Open cdbethune opened 4 years ago
The JIDO dataset has a tabular version that is being used for the eval, so we don't need to do anything with this yet.
The dataset driving this use case will be handled by TA2 systems on their own, so while we should be able to display a text view, this is no longer a priority.
We need to support full display of text documents. Right now for text classifcation problems, we just display a link in the table/facets. At minimum we need to allow the user to click on the document as a link in the table, and bring up a modal that has the complete text in it. Better would be some sort of an analog of the image mosaic view, where we bring up a document-centric view, although that will require design.