This plugin offers function score conditional boosting, depending on values of other fields in a document.
Note, conditional boosting is also possible by scripting.
This plugin demonstrates how to implement a solution without scripting by extending the Elasticsearch function score module.
Imagine a shop wants to advertise non-sellers in the search list of the online shop without telling the visitors. Your boss comes in and in a harsh voice you hear the command, "Push those damned non-sellers out this month and keep the search response time low, or you are fired!"
Your idea is to use document boosting for changing the hit orders of the shop search lists. Normally, you would have to re-index all the content with adjusted boost values in a boost field, or use a script for function score, to mangle with the document fields, evaluating them, and change the boost.
With this plugin, you can fake the relevance ranking on the fly, without reindexing, but also without scripting.
Assume the product product_name_1 has to appear in the top results, you can formulate
a boost condition with a specific boost value for product_name_1, and the score value will be modified
for these documents.
Assume also a supplier user_name_1 who has paid you an illegal extra bonus if you rank him
high. You could combine these boosts, many string field and string values can be specified
to trigger a combined conditional document boost in a single query.
If there is more than one match in the specified field values in a document, scores multiply each other.
For smoothing the scoring function, you can use a factor and a modifier like
in field value factor score function, see here
Full example:
curl -XDELETE '0:9200/test'
curl -XPUT '0:9200/test' -d '
{
"index.number_of_shards" : 2,
"index.number_of_replicas" : 1
}
'
curl -XPUT '0:9200/test/products/1' -d '
{
"content" : "foo bar",
"product" : "product_name_1"
}
'
curl -XPUT '0:9200/test/products/2' -d '
{
"content" : "foo bar",
"product" : "product_name_2"
}
'
curl -XPUT '0:9200/test/products/3' -d '
{
"content" : "foo bar",
"product" : "product_name_3"
}
'
curl -XPUT '0:9200/test/products/4' -d '
{
"content" : "foo bar",
"product" : "product_name_1",
"user" : "user_name_1"
}
'
curl -XPUT '0:9200/test/products/5' -d '
{
"content" : "foo bar",
"product" : "product_name_2",
"user" : "user_name_1"
}
'
curl -XPUT '0:9200/test/products/6' -d '
{
"content" : "foo bar",
"product" : "product_name_3",
"user" : "user_name_1"
}
'
curl -XGET '0:9200/test/_refresh'
curl -XGET '0:9200/test/products/_search' -d '
{
"query":{
"function_score":{
"query":{"term": { "content" : "foo"} },
"cond_boost":{
"cond":[
{"product":"product_name_1","value":2.0},
{"user":"user_name_1","value":3.0}
],
"value":1.0,
"factor":1.0,
"modifier":"NONE"
}
}
},
"explain":false
}
'
| Plugin | Elasticsearch version | Release date |
|---|---|---|
| 2.3.3.0 | 2.3.3 | Jun 19, 2016 |
| 1.4.0.0 | 1.4.0 | Feb 19, 2015 |
| 1.2.1.0 | 1.2.1 | Jun 7, 2014 |
./bin/plugin install http://xbib.org/repository/org/xbib/elasticsearch/plugin/elasticsearch-condboost/2.3.3.0/elasticsearch-condboost-2.3.3.0-plugin.zip
./bin/plugin -install condboost -url http://xbib.org/repository/org/xbib/elasticsearch/plugin/elasticsearch-condboost/1.4.0.0/elasticsearch-condboost-1.4.0.0.zip
Do not forget to restart the node after installing.
All feedback is welcome! If you find issues, please post them at Github
Elasticsearch Conditional Boost Plugin
Copyright (C) 2014 Jörg Prante
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.