This PR enhances the dataset card by adding:
- Correct task category (`other`) and relevant tags (time-series, natural-language-querying, symbolization)
- Links to the paper and GitHub repository
- A brief description of the dataset
- A sample usage code snippet taken from the official GitHub README
Remove 1153 misplaced root-level CSVs (kept under ts_data/)