[Edit]

TypeScript - check words similarity (fuzzy compare with bigrams)

1 contributors

2 contributions

0 discussions

0 points

Created by:

cory

1756

In this article, we would like to show how to check words similarity in TypeScript.

Below logic:

calculates words bigrams,
counts bigram hits to find similarity,
divides hits by bigrams to calculate final words similarity.

Below checkSimilarity() function result indicates how two words are similar.

Similarity measured is from 0 where:

0 - means: the worlds are totally different,
>=1 - means: the words are the same or contain similar part.

That kind of approach is may be applied in fuzzy search.

Practical example:

xxxxxxxxxx
 
// ONLINE-RUNNER:browser;
​
const createBigram = (word: string): string[] => {
  const input = word.toLowerCase();
  const vector = [];
  for (let i = 0; i < input.length; ++i) {
    vector.push(input.slice(i, i + 2));
  }
  return vector;
};
​
const checkSimilarity = (a: string, b: string): number => {
  if (a.length > 0 && b.length > 0) {
    const aBigram = createBigram(a);
    const bBigram = createBigram(b);
    let hits = 0;
    for (let x = 0; x < aBigram.length; ++x) {
      for (let y = 0; y < bBigram.length; ++y) {
        if (aBigram[x] === bBigram[y]) {
          hits += 1;
        }
      }
    }
    if (hits > 0) {
      const union = aBigram.length + bBigram.length;
      return (2.0 * hits) / union;
    }
  }
  return 0;
};
​
​
// Usage example:
​
console.log(checkSimilarity('Chris', 'Chris'));  // 1
console.log(checkSimilarity('John1', 'John2'));  // 0.6
console.log(checkSimilarity('Google', 'Gogle')); // 0.9090909090909091
console.log(checkSimilarity('Ann', 'Matt'));     // 0