HomeData ScienceRecommender Popularity Bias (2D): The Rich-Get-Richer Feedback Loop

Recommender Popularity Bias (2D): The Rich-Get-Richer Feedback Loop

Interactive 2D simulator of the Matthew effect in recommendation systems: a radial exposure chart, a live Lorenz curve and a concentration-over-time graph show collaborative-filtering feedback loops concentrate exposure onto a shrinking set of items as bias strength, exploration rate and list size change.

Data Science2DModerate60 FPS📱 Mobile-adapted⇄ 3D version
2d-ds-topic-35 ↗ Open standalone

Recommendation systems don't just reflect what users like — the act of recommending changes what gets seen, which changes what gets interacted with, which changes what gets recommended next. This 2D simulator makes that closed loop visible across three linked panels: a radial exposure chart where 72 catalogue items sit as bars around a ring, a live Lorenz curve quantifying inequality, and a rolling history of Gini coefficient and top-10% share. Push the popularity bias α up and watch a handful of bars pull away into a spike while the Lorenz curve sags toward the bottom-right corner; raise the exploration rate ε and watch both flatten back out. Drag the radial chart to rotate your view of the catalogue while the feedback loop runs.

⚙ Under the hood

Watch collaborative-filtering exposure feedback concentrate recommendations onto a shrinking set of items across three linked 2D panels — a radial exposure chart, a live Lorenz curve and a concentration-over-time graph. Tune popularity bias strength, exploration rate and list size, and track the Gini coefficient, top-10% exposure share and long-tail percentage live.

recommendation systemscollaborative filteringpopularity biasMatthew effectGini coefficientLorenz curvedata science

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install

What did you find?

Add reproduction steps (optional)