Fuzzy Search Algorithms How and When to Use Them
summary
IPython notebook walks through concrete implementations: uses Jellyfish library for Soundex and Levenshtein, queries a local PostgreSQL instance with the PostgreSQL trigram extension, and runs similarity scoring with Jaro and trigram Jaccard. Demo data includes 5,000 Hillary emails and a small misspelled-city table where Soundex returns three Washington DC hits but trigram finds four. Shows Levenshtein quirks (99 vs 100 Broadway) and that Damerau handles transpositions; shows word similarity via word Toc vectors and cosine distances.
AI-written summary · about
chapters (10)
- 0:00Fuzzy search algorithms intro
- 2:10Soundex explained
- 8:54Soundex in PostgreSQL queries
- 11:44Levenshtein distance basics
- 16:42Levenshtein with transpositions
- 17:24Trigram similarity matching
- 20:39PostgreSQL trigram demo
- 24:29Semantic similarity embeddings
- 27:50Model integration tips
- 31:27Q&A
related talks
42mYou don't need Elasticsearch! Fuzzy Search with PostgreSQL and Spring Data
24mWhat's in a name? Phonetic Algorithms for Search and Similarity
13mUnconference Session: A Neat Trick for Fuzzy Search
43mFull-Text-Search Explained from A to Z with French Food
26mOn The Look-Out For Your Data
30mFuzzy Matching at Scale
31mThe drunk Gopher’s walk: A talk about fuzzy, I Ching and ale.
49mBreakout Session