Hi Erick,
Yeah, I've reviewed the debug output and can't make sense of why they
are scoring the same. I have double checked that they are being indexed
with different boost values for the search field. I've also increased
the factors trying to get them be more granular so instead of boosting
1,2,3,4,5 I did 100,200,300,400,500... Same result.
Here's and example of the debug output with two documents having
different field boost values but receiving the same score.
Does anything stick out? Any other ideas on how to get the results I am
looking for?
69.694855 = (MATCH) product of: 104.54228 = (MATCH) sum of: 0.08869071 =
(MATCH) MatchAllDocsQuery, product of: 0.08869071 = queryNorm 104.45359
= (MATCH) weight(searchfe2684d248eab25404c3668711d4642e_boost:true in
4016) [DefaultSimilarity], result of: 104.45359 =
score(doc=4016,freq=1.0 = termFreq=1 ), product of: 0.48125002 =
queryWeight, product of: 5.4261603 = idf(docFreq=81, maxDocs=6856)
0.08869071 = queryNorm 217.04642 = fieldWeight in 4016, product of: 1.0
= tf(freq=1.0), with freq of: 1.0 = termFreq=1 5.4261603 =
idf(docFreq=81, maxDocs=6856) 40.0 = fieldNorm(doc=4016) 0.6666667 =
coord(2/3)
69.694855 = (MATCH) product of: 104.54228 = (MATCH) sum of: 0.08869071 =
(MATCH) MatchAllDocsQuery, product of: 0.08869071 = queryNorm 104.45359
= (MATCH) weight(searchfe2684d248eab25404c3668711d4642e_boost:true in
4106) [DefaultSimilarity], result of: 104.45359 =
score(doc=4106,freq=1.0 = termFreq=1 ), product of: 0.48125002 =
queryWeight, product of: 5.4261603 = idf(docFreq=81, maxDocs=6856)
0.08869071 = queryNorm 217.04642 = fieldWeight in 4106, product of: 1.0
= tf(freq=1.0), with freq of: 1.0 = termFreq=1 5.4261603 =
idf(docFreq=81, maxDocs=6856) 40.0 = fieldNorm(doc=4106) 0.6666667 =
coord(2/3)
On 1/11/2012 9:46 PM, Erick Erickson wrote:
Boosts are fairly coarse-grained. I suspect your boost factors are just
being rounded into the same buckets. Attaching&debugQuery=on and
looking at how the scores were calculated should help you figure out
if this is the case.
Best
Erick
On Wed, Jan 11, 2012 at 7:57 PM, Brett<br...@chopshop.org> wrote:
I'm implementing a feature where admins have the ability to control the
order of the results by adding a boost to any specific search.
The search is a faceted interface (no text input) and which we take a hash
of the search parameters (to form a unique search id) and then boost that
field for the document.
The field is a wild card field so that it might look like this:
<field name="search395eff966b26a91d82935c8e1197330c_boost"
boost="90">true</field>
The problem is that in these search results I am seeing is that my results
are being grouped and the individual boost values are not having the
granular effect I am looking for.
Say on a result set of 75 documents. I see results with search boosts of
60-70 receiving the same score even though they were indexed with different
boost values. There are always more than one group.
Does anyone know what might be causing this? Is there a better way to do
what I am looking for?
Thanks,
Brett
Field Definition:
<fieldType name="boost" class="solr.TextField" sortMissingLast="true"
omitNorms="false" omitTermFreqAndPositions="true">
<analyzer type="index">
<tokenizer class="solr.WhitespaceTokenizerFactory"/>
<filter class="solr.LowerCaseFilterFactory"/>
</analyzer>
<analyzer type="query">
<tokenizer class="solr.WhitespaceTokenizerFactory"/>
<filter class="solr.LowerCaseFilterFactory"/>
</analyzer>
</fieldType>
