#community-help

Calculating Text Match Score as a Percentage

TLDR Rohit asked how to determine a percentage match for text searches. Kishore Nallan explained the computation process and shared about an upcoming feature that will provide more detailed statistics.

25mo

Join the chat

Dec 15, 2021 (25 months ago)

Photo of md5-35b1799bd10ed2978c0e2c07bdfa0aa5

Rohit

07:55 AM

For example, If a document has text : Hello World! and I search for Hello . How can I know its a 45-50% match?

Photo of md5-4e872368b2b2668460205b409e95c2ea

Kishore Nallan

09:46 AM

👋 Rohit The text match score is an amalgamation of various components like number of words matching, number of typos, distance between matches etc. So there is no way to treat it as a percentage match.

09:47

Kishore Nallan

09:47 AM

We do have an item in the backlog to break these individual score components down so when we do that, you will be able to pull out finer grain statistics

Rohit

09:51 AM

Kishore Nallan That is great. Thanks for the update. Typesense is really great. I'm doing a small POC for some project and this has helped me.

Kishore Nallan

09:53 AM

Glad to hear!

Typesense

Lightning-fast, open source search engine for everyone | Knowledge Base powered by Struct.AI

Indexed 3015 threads (79% resolved)

Join Our Community

Similar Threads

Discussing 'Text Matches' Feature Improvement in Typesense

Stefan asked if distance in 'text matches' can be determined. Kishore Nallan said no and asked for an issue to be created on Github, which was done by Stefan.

27mo

Text Match Score Calibration for Near Exact Matches

Richard asked about calibrating text match score for near exact matches. Jason explained its purpose and suggested parameters for exact matches, referring to Typesense documentation.

4mo

Understanding the 'text_match' calculation in Typesense.

John asked about the interpretation of `text_match` scores while Kishore Nallan clarified that the score is a composed int64 integer and shared the computation link.

18mo