Depending on your specific use case, you can handle this differently. You can give those items a low rating in your ranking stage, or re-recommend them with high probablity (for example, you want your user to resume watching a video or placing an order on a previously seen item), or just filtering them out. You can also make it a feature.
Nice!
Thank You
How to deal with items that the user has already seen/bought/interacted with?
Depending on your specific use case, you can handle this differently. You can give those items a low rating in your ranking stage, or re-recommend them with high probablity (for example, you want your user to resume watching a video or placing an order on a previously seen item), or just filtering them out. You can also make it a feature.
Pip can not find tensorflow-quantum
Is there any progress on making the current TensorFlow version more compatible with Keras?