Find the closest matching strings in a candidate list with multiple fuzzy matching algorithms
Zero Dependencies Β· Works Offlineπ Search Term
π Candidate List (one per line)
What Can the Fuzzy String Matcher Do?
The Fuzzy String Matcher is a free online tool that quickly finds the most similar strings to your search term in a candidate list. It supports four classic fuzzy matching algorithms β Soundex, Metaphone, N-gram, and Levenshtein β with side-by-side comparison. Ideal for spell correction, name matching, and data deduplication. All matching runs locally in your browser with no server uploads.
Core Features
- Multi-algorithm Comparison: Soundex phonetic encoding, Metaphone voice matching, N-gram substring similarity, Levenshtein edit distance
- Real-time Search: Results update as you type, sorted by similarity score
- Custom Candidate Lists: Paste multi-line text as candidates
- Adjustable Parameters: N-gram size, max results count
- Multi-format Export: Copy results or download as CSV
How to Use
1. Enter the string you want to search for.
2. Enter candidate words in the candidate area, one per line.
3. Select the matching algorithm and parameters.
4. Click "Search" or results update automatically as you type.
5. View results sorted by similarity score.
Use Cases
Case 1: Spell Correction
Use Levenshtein distance to find the closest correct word in a dictionary for misspelled input.
Case 2: Name Matching
Use Soundex or Metaphone to match phonetically similar names, solving variants like "Smith" vs "Smyth".
Case 3: Data Deduplication
Use N-gram similarity to discover duplicate records with slight spelling differences during data cleaning.
Case 4: Autocomplete
Implement fuzzy autocomplete in search boxes that finds relevant results even with imprecise input.
Extended Knowledge
Fuzzy string matching (Approximate String Matching) is a classic problem in computer science. Soundex, invented by Robert Russell and Margaret Odell in 1918, was the first phonetic encoding algorithm, mapping words to 4-character codes (initial letter + 3 digits). Metaphone, proposed by Lawrence Phillips in 1990, improved on Soundex's English pronunciation rules. Double Metaphone further supports multiple language pronunciation variants. N-gram methods split strings into substrings of length N and measure similarity by comparing substring set overlap β they don't depend on pronunciation rules, making them suitable for non-English text. Levenshtein distance measures edit operation count and is the most intuitive similarity metric.
β FAQ
What algorithms are supported?
Soundex, Metaphone, N-gram similarity, and Levenshtein distance β all comparable side by side.
Soundex vs Metaphone?
Soundex is a simple 4-character phonetic code. Metaphone is more precise with better English pronunciation rules. Metaphone is better for name matching.
Does it support non-English text?
Soundex and Metaphone are designed for English. For non-English, use N-gram or Levenshtein which work on character comparison.
Is my data uploaded?
No. Pure frontend technology β all matching happens locally. Your data never leaves your device.
Can I customize the candidate list?
Yes. Enter multiple lines of text, one candidate per line.
Can I export results?
Yes. Copy results to clipboard or download as CSV file.