: Boost Query (bq) accepts lucene queries. E.g. bq=price:[50 TO 100]^100
: boost and bf parameters accept Function queries, e.g. boost=log(popularity)
while these statements are both true, they don't tell the full story.
for example you can also specify a function as a query using the
appropriate parser: bq={!func}log(popularity)
or turn any query into a function that produces values according to the
query score: boost=query({!lucene v='price:[50 TO 100]^100'})
The fundemental difference between bq & boost is:
"bq" causes a an additional 'boost query' clause to be *added* to your
original query
"boost" causes the scores for each doc from your original to
be *multiplied* by the results of the specified function evaluated
against the same doc.
(in both cases "original query" refers to your "q" param
parsed with respects to qf, pf, etc...)
So a query like this...
q={!edismax}bar & qf=foo & bq=x:y
...is roughly equivilent to: q={!lucene}+foo:bar x:y
While a query like this...
q={!edismax}bar & qf=foo & boost=query({!lucene v='x:y'})
...is roughly equivilent to...
q={!func}prod(query({!edismax qf='foo' v='bar'}), query({!lucene v='x:y'}))
because of how they affect final scores, the 'boost' param is almost
always what you really want and is really nothing more then shorthand for
wrapping your entire query in a "BoostQParser" ...
https://cwiki.apache.org/confluence/display/solr/Other+Parsers#OtherParsers-BoostQueryParser
-Hoss
http://www.lucidworks.com/
