Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Frontend Fuzzy Search

github.com

Post reply on HN
Print view View original

1–10 of 38 posts

Show HN: Frontend Fuzzy Search

#1

Post by kmschaal »

I have spent several years working on search engines in the backend and have now utilized that experience to develop a fuzzy search library for the frontend. It's fast, accurate and can be used for all languages. It should be easy to integrate into your Javascript / Typescript projects. If you test it and find any edge cases that did not work for you please let me know.

The implementation is based on 3-grams by the book, augmented with a novel trick of sorting the characters within the 3-grams for enhanced accuracy. For a detailed explanation you may refer to my related blog post at https://www.m31coding.com/blog/fuzzy-search.html.

Happy Coding!

Show HN: Frontend Fuzzy Search
github.com

Re: Show HN: Frontend Fuzzy Search

#5

Post by kmschaal »

tkcranny wrote: post #3

Looks like a neat library. I’m curious if there’s a tl;dr on how this is better or different to Fuse, which is a very popular established client side fuzzy searching library. https://github.com/krisk/fuse

Hi, thank you for your interest! I believe that most fuzzy search implementations lack accuracy in one aspect or another. The primary goals of my library are accuracy and query performance. However, I haven't looked into Fuse yet. I'm highly interested in hearing feedback from people who have tried both libraries with their datasets.

Re: Show HN: Frontend Fuzzy Search

#7

Post by swyx »

tkcranny wrote: post #3

Looks like a neat library. I’m curious if there’s a tl;dr on how this is better or different to Fuse, which is a very popular established client side fuzzy searching library. https://github.com/krisk/fuse

for those looking for options i found ufuzzy a lightweight and good enough version of fuse: https://swyxkit.netlify.app/ufuzzy-search

Re: Show HN: Frontend Fuzzy Search

#8

Post by LoganDark »

kmschaal wrote: post #5
tkcranny wrote: post #3

Looks like a neat library. I’m curious if there’s a tl;dr on how this is better or different to Fuse, which is a very popular established client side fuzzy searching library. https://github.com/krisk/fuse

Hi, thank you for your interest! I believe that most fuzzy search implementations lack accuracy in one aspect or another. The primary goals of my library are accuracy and query performance. However, I haven't looked into Fuse yet. I'm highly interested in hearing feedback from people who have tried both libraries with their datasets.

What is your definition of "accuracy" within the context of fuzzy search?

Re: Show HN: Frontend Fuzzy Search

#9

Post by kmschaal »

LoganDark wrote: post #8
kmschaal wrote: post #5

Earlier quoted context omitted.

Hi, thank you for your interest! I believe that most fuzzy search implementations lack accuracy in one aspect or another. The primary goals of my library are accuracy and query performance. However, I haven't looked into Fuse yet. I'm highly interested in hearing feedback from people who have tried both libraries with their datasets.

What is your definition of "accuracy" within the context of fuzzy search?

It's subjective, I have to admit. I would say a search is accurate if most people find what they are looking for in their dataset in the first try.

Distance definitions such as the Levenshtein and Damerau-Levenshtein distances provide a solid basis for discussions on accuracy. However, they are costly to compute and hence not widely adopted in fuzzy search libraries.

I started by using the known filter equation for the Levenshtein distance and computed a quality score with a leightweight formula. Then, I realized that the filter equation can be extended to the Damerau-Levenshtein distance by sorting the characters of the 3-grams.

In my tests, this implementation worked well. Please let me know how it works for you if you test it.

Post reply on HN