bmpm-phonetics

TypeScript implementation of the Beider–Morse Phonetic Matching (BMPM) algorithm.
This library encodes personal names into phonetic keys across multiple languages,
allowing approximate or cross-linguistic matches such as:

| Name A | Name B | Match | |-------------|-------------|:-----:| | Schmidt | Smith | ✅ | | Schwarz | Shwartz | ✅ | | Rodrigues | Rodriguez | ✅ | | Cohen | Kohn | ✅ | | Levi | לוי | ✅ | | Nowak | Novak | ✅ |

Features

• Full Beider–Morse algorithm implemented in TypeScript
• Multi-language rule sets (English, German, French, Spanish, Portuguese, Polish, Russian, Hebrew)
• Built-in transliteration for Cyrillic and Hebrew scripts
• Exact and Approximate matching modes
• Returns both boolean matches and similarity scores
• Suitable for data cleaning, deduplication, genealogy, and record linkage

Installation

npm install bmpm-phonetics

or

yarn add bmpm-phonetics

Quick Start

import {
  bmpmMatch,
  bmpmSimilarity,
  ExtendedBMPMConfig as cfg
} from "bmpm-phonetics";

console.log(bmpmMatch("Schmidt", "Smith", cfg));
// → true

console.log(bmpmSimilarity("Cohen", "Kohn", cfg));
// → 0.46