On 02/14/2017 07:35 AM, Thomas Nyberg wrote:
> Hello,
>
> I think it's easier to explain my question with example code:
>
> ------------------------
> CREATE TABLE t ( s VARCHAR );
> CREATE TABLE
>
> INSERT INTO t VALUES ('hello'), ('hello world');
> INSERT 0 2
>
> SELECT * FROM t;
> s
> -------------
> hello
> hello world
> (2 rows)
>
> SELECT s, ts_rank(vector, query) AS rank
> FROM t, to_tsvector(s) vector, to_tsquery('hello') query
> WHERE query @@ vector;
> s | rank
> -------------+-----------
> hello | 0.0607927
> hello world | 0.0607927
> (2 rows)
> ------------------------
>
> Here both 'hello' and 'hello world' are ranked equally highly when
> searching with 'hello'. What I'm wondering is, is there a way within
> postgres to have it match higher to just 'hello' than 'hello world'?
> I.e. something like it slightly down-weights extraneous terms? Of course
> in general I don't know the query or the field strings ahead of time.
Some digging around found this:
https://www.postgresql.org/docs/9.6/static/textsearch-controls.html#TEXTSEARCH-RANKING
Setting a normalization of 1:
test=# SELECT s, ts_rank(vector, query, 1) AS rank
FROM t, to_tsvector(s) vector, to_tsquery('hello') query
WHERE query @@ vector;
s | rank
-------------+-----------
hello | 0.0607927
hello world | 0.0383559
>
> Thanks for any help!
>
> Cheers,
> Thomas
>
>
--
Adrian Klaver
adrian.klaver@xxxxxxxxxxx
--
Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
