[Question] Possibility to extend to images for a multi-modal dataset

[m3at]

Hello and thank you for this nice dataset.

In the data, it seems that the product_id field map to Amazon's ASIN numbers. If this is correct, we might be able to retrieve items (example using the product advertising api) and associated images, which would make for an interesting multi-modal dataset.
Is this something you are considering?

[franbvalero]

Hello @m3at , thanks a lot for your question and the interest in the dataset.

Yes, the product_id is the ASIN of the product, so you can use it to retrieve the associated images.

For this first published release version of the dataset we did not consider it. We had it in mind, but we did not include it, because we wanted to build a text dataset.

[m3at]

Fair enough, thanks for the reply 👍