🛍️
The left panel shows a small e-commerce taste space. Products are styled icons, and each consumer
is a boy or girl avatar. Consumers give higher ratings to nearby products. A product’s colored halo shows its
k-center style preference radius.
🛒 Products and consumers in a 2D preference space
Products (P0, P1, …) are shown as larger “cards” in the middle of the board. Consumers (C0, C1, …) appear as
boy/girl avatars around them. Each consumer is assigned to the product with the highest rating (closest center).
The halo around each product reaches its farthest consumer.
🖥️ product card (Pj) with preference halo
👦 boy consumer (C i)
👧 girl consumer (C i)
Tip: click a product card to list its consumers on the right.
Global max preference radius
–
Average radius across products
–
Coverage comment
–
🎯 Selected product and its consumers
Click a product icon on the left. This table shows all boy/girl consumers inside that product’s preference radius.
The radius is measured as Euclidean distance in the 2D taste space.
Product
Consumer
Gender
Distance
Rating
Tag
📊 Sample rating matrix (first 8 consumers)
Each entry is a 1–5 score based on distance: closer consumers tend to give higher scores. The assigned product
is the one with the highest rating (ties broken randomly).
📜 Event log
This log connects steps in the simulation to ideas from k-center, clustering, and recommender systems.