Greetings, dear Community.
It would be great if somebody can help me.
We have n-types in index, every type has own set of columns(fields).
When I want to search above all index, I need to configure relevance depending on type and type field.
How can I do it?
Thank you for your attention.
What do you mean by "configure relevance depending on type and type field"?
Do you want to customize TF*IDF per type/field? If so, follow these instructions and change the similarity per mapping.
Do you want to control the matching process by customizing analyzers? Well you can do that too.
Or do you want to control how different fields are weighted/boosted. Well there's that too.
But that covers basically an extremely broad range of capabilities, so I'm not sure what you mean. There's also a whole book (disclaimer: that I wrote) that covers relevance in Elasticsearch you might want to check out. DM me for a discount code.
1 Like
I mean, I have index with two types like "Contacts", and "Account", they both have "Name" field, how can I boost
field "Name" in "Contact" higher than "Name" in "Account"?