Document comparison / similarity using Whoosh Python Search Library
Question:
How do I get a similarity measure of a document using Whoosh?
I want to create a “Related” feature that ranks other previously indexed documents that have a high similarity to a document.
Do I input the document as a long query string? Do I add the document to the index and extract a similarity query result somehow from there?
Thanks
Answers:
The Whoosh searcher class has a method called ‘more_like()’.
It allows you to compare and indexed document to other indexed documents and returns a list of documents similar to the given document.
And the class whoosh.searching.Hit can give a rank and a score.
Updated links:
more_like() : https://whoosh.readthedocs.io/en/latest/api/searching.html#whoosh.searching.Searcher.more_like
whoosh.searching.Hit : https://whoosh.readthedocs.io/en/latest/api/searching.html#whoosh.searching.Hit
How do I get a similarity measure of a document using Whoosh?
I want to create a “Related” feature that ranks other previously indexed documents that have a high similarity to a document.
Do I input the document as a long query string? Do I add the document to the index and extract a similarity query result somehow from there?
Thanks
The Whoosh searcher class has a method called ‘more_like()’.
It allows you to compare and indexed document to other indexed documents and returns a list of documents similar to the given document.
And the class whoosh.searching.Hit can give a rank and a score.
Updated links:
more_like() : https://whoosh.readthedocs.io/en/latest/api/searching.html#whoosh.searching.Searcher.more_like
whoosh.searching.Hit : https://whoosh.readthedocs.io/en/latest/api/searching.html#whoosh.searching.Hit