How to use Similarity
- 1. Paste string A and B. Case-sensitive for Levenshtein; Dice lowercases.
- 2. Read distances and scores. Normalized Levenshtein is 1 - d/maxLen.
- 3. Copy the JSON report. Use it in a test fixture.
- 4. No ML. These are classic metrics.
About this tool
Fuzzy matching starts with distances. Levenshtein counts edits. Jaro rewards matching characters within a window. Dice uses bigram overlap. This tool reports all three so you can pick a threshold for deduping names or logs.
Scale
Very long strings allocate an (n+1)×(m+1) table. Keep inputs reasonable.
Dice
Single-character strings have no bigrams; identical empty strings score 1.
Code examples
Pair
kitten
sitting