🎁 Intra-List Diversity (ILD)

Intra-List Diversity (ILD)#

The average dissimilarity among items within a single recommendation list.

Important

✨ AI-generated content. This page was written with the assistance of an AI language model and is provided as a learning aid. Despite careful review, it may still contain mistakes, omissions, or out-of-date information. Whether you are new to the topic, a team lead, or a senior practitioner, treat it as a starting point rather than an authoritative reference: read it critically and independently verify anything you act on (code, commands, figures, and factual claims) against official documentation and primary sources before relying on it.

What it is#

Intra-list diversity measures how varied the items within a single recommendation list are — the antidote to lists of near-identical products. It is defined from the intra-list similarity (ILS), the average pairwise similarity of all items in the list; diversity is its complement.

The formula#

\[\mathrm{ILD} = \frac{2}{|L|\,(|L|-1)} \sum_{i < j} \big(1 - \mathrm{sim}(i, j)\big),\]

with \(\mathrm{sim}\) a cosine (over embeddings / features) or Jaccard (over tags / genres) similarity. High ILS → similar items → low diversity; low ILS → varied items → high diversity.

Why it matters#

Accuracy alone rewards recommending ten versions of the same hit; diversity captures whether a list actually broadens what the user sees. It trades off against relevance — the art is a diverse yet relevant list.

Theme: Recommender Systems · All terminology

Hint

Mind map — connected ideas

Hint

More in Recommender Systems

Catalog Coverage · Cosine Similarity of Item Features · Diminishing Utility · Diversity (in Recommender Systems) · Dominating in Recommender Systems · Genre Overlap · Hit Rate (HR) · Item Coverage · Jaccard index · Novelty (in Recommender Systems) · Relevance in Recommender Systems · Self-Information of Popularity · User Coverage