Hi Dave,
That's possible in 3.5, check out the following example:
*FOR d in myView *
*SEARCH d.name IN [ 'foo', 'bar', 'baz' ] *
*SORT BM25(d)*d.rating DESC // DESC is important to get the most relevant
results first*
*RETURN d*
you can return evaluated value as well:
*FOR d in myView *
*SEARCH d.name IN [ 'foo', 'bar', 'baz' ] *
*LET score = BM25(d)*LOG(d.rating+1)*
*SORT score DESC*
*RETURN { d, score }*
you can also use query time boost to modify overall score based on your
search condition:
LET importance = 42
*FOR d in myView *
*SEARCH BOOST(**d.name == 'importantName', importance) || **d.name IN [
'foo', 'bar', 'baz' ] *
*LET score = BM25(d)*LOG(d.rating+1)*
*SORT score DESC*
*RETURN { d, score }*
On Thursday, 30 May 2019 13:02:50 UTC+3, Dave Challis wrote:
>
> For example, say I've got collection of products containing data such as:
>
> {"name": "headphones", "rating": 4}
>
> Is there any way (using a fulltext index, or search view, or custom
> function) to boost/modify the searches on the "name" field based on the
> "rating" field?
>
> I'd like to allow the most relevant products to be found, but then also
> apply a small upwards boost for things with a high rating field.
>
--
You received this message because you are subscribed to the Google Groups
"ArangoDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/arangodb/2c9a995d-4d24-4ad2-a4b8-5b7129933720%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
