Open Bakha87 opened 1 year ago
Pinging @elastic/es-search (Team:Search)
Jaro-Winkler focuses on the edit distance of strings, similar to Levenshtein. Elasticsearch does text search via an inverted index. Scoring documents based on tokens/terms matched via some algorithm (default being BM25).
I don't see how J-W would work in an inverted index system.
Can you expound on how are you expecting to use this during search and why do you need it?
I need to select all records that match the search string by more than 85%. I can give an example of a request in Oracle
Pinging @elastic/es-search-relevance (Team:Search Relevance)
Description
Could you please add searching using Jaro–Winkler similarity algorithm feature in next releases or provide a plugin that implements the mentioned functionality. Thank you