Jaccard Similarity
Definition
Jaccard similarity measures, for a non-empty union of two sets, the size of their intersection divided by the size of their union.
For a non-empty union, the value ranges between 0 and 1. For two empty sets, a convention must be established. The measure is suitable for binary or set-like data and, unlike some other similarity measures, does not consider jointly missing features.
Simply explained
How to compare how many elements two lists have in common – relative to all distinct elements combined.
Why it matters in practice
Relevant for deduplication, tag comparisons, cluster analysis, and certain retrieval tasks.
Example from business
Two product datasets are checked for possible duplicates based on their normalized feature sets.
Limits and pitfalls
- Applying Jaccard to continuous vectors without validation; not defining empty sets; setting thresholds without evaluation.
Origin & history
Libraries like scikit-learn support Jaccard evaluations for set-like or binary classification data.
Use in mid-sized companies
- Possible applications: similarity of tags, sets, document features, or binary attributes.
Frequently asked questions
Is Jaccard the same as Cosine Similarity?
No. What does 0 mean? For a non-empty union, no common set overlap. For two empty sets, a convention must be established.
Sources
- jaccard_score · scikit-learn, jaccard_score
Sources checked on 10.10.2026.
Editorially reviewed on: