Multiple Dataset management

Dataset as a tree/graph. If you can imagine that a dataset will undergo changes like normalization and/or context changes, it needs to be presented as such with transparency for the user and agents. Also it can be across different domains (such as how cs-analyze would generate new datasets, it can then be imported into this management system. ) so a selection system is needed, since this lean ecoxai is only for a single dataset

A dataset should be considered one or more ingestible data. the user context/research question should be on that dataset, however complex

Current implementation:
single dataset upload (doesn't handle multiple files for a single dataset yet)
single folder (/backend/datasets)

Future features:
multiple datasets selection (updates in the sql and workflow for data and datasets folder needed)
backend assets management
versioning
integration into the hypothesis graph as bonus

Start by reading the current single-dataset upload flow and the /backend/datasets folder. Then inspect the SQL and workflow paths mentioned in the issue, including the data and datasets folders. Done requires a defined approach for selecting multiple datasets and updating the related storage and workflows; versioning, asset management, and hypothesis-graph integration are listed as future scope.

Written by the indexing model from the issue text.