Shopping Basket Simulator
Simulate a customer's basket and see what products the system would recommend — blending KNN similarity and association rule confidence.
Add products to basket
Loading products...
Current basket (0 items)
No items in basket yet. Click products on the left to add them.
Recommended products
Add products to the basket, then click "Get recommendations" to see what a customer would likely also buy.
How recommendations are computed
1. KNN similarity
For each item in the basket, look up its K nearest neighbours (cosine similarity on co-occurrence vectors). Sum the scores across all basket items.
2. Association boost
Boost products that have strong association rules (high confidence) from basket items. This catches directional affinities KNN alone might miss.
3. Rank & deduplicate
Sort by combined score, remove items already in the basket, and return the top-N most likely additions.