Elasticsearch Normalised Score with Boost Documents - elasticsearch

Question

I am building a query that takes a set of codes and geo_point locations. The result should be a list of documents ordered by distance to origin. However, I would like to be normalised with let say a score of 10 for the document in the origin location and decreasing according to distance from origin. I have actually managed to build this search but I also would like to increase the score of a document if this includes an additional variable in the list of codes.
These are the requirements:
The output should be a list of documents which score is normalised according to distance from origin.
Documents returned should contain at least one yvar (i.e. yvar1 OR yvar2 OR yvar3 OR yvar...).
Only documents after a certain date should be returned
Only documents containing all the xvars passed to the query must be returned.
If a document has an additional x variable (e.g xvar4) the score for this document, should be increased by 0.1. This is the bit I am struggling with.
This is my mapping:
{
"mappings": {
"properties": {
"codes": {
"type": "keyword"
},
"date": {
"type": "date",
"format": "dd/MM/yyyy"
},
"coordinates": {"type": "geo_point"}
}
}
}
Some example documents (NB: The distanceToOrigin is for analysing the output only):
{ "create" : { "_index": "my-index", "_id" : "1" } }
{ "id": 1, "coordinates": { "lat": 51.5132, "lon": -0.1362}, "available capacity": 5, "last updated": "01/11/2021", "ResponseCodes": ["xvar1", "xvar2", "xvar3", "yvar1", "yvar2", "yvar3" ] ,"distanceTOorigin": 0 }
{ "create" : { "_index": "my-index", "_id" : "2" } }
{ "id": 2, "coordinates": { "lat": 52.9114, "lon": 0.5580}, "available capacity": 5, "last updated": "01/11/2021", "ResponseCodes": ["xvar1", "xvar2", "xvar3", "xvar4", "yvar1", "yvar2", "yvar3" ] ,"distanceTOorigin": 114 }
{ "create" : { "_index": "my-index", "_id" : "3" } }
{ "id": 3, "coordinates": { "lat": 51.4890, "lon": -0.6029}, "available capacity": 5, "last updated": "01/11/2021", "ResponseCodes": ["xvar1", "xvar2", "xvar3", "yvar1", "yvar2", "yvar3" ] ,"distanceTOorigin": 22 }
{ "create" : { "_index": "my-index", "_id" : "4" } }
{ "id": 4, "coordinates": { "lat": 57.2555, "lon": -3.2692}, "available capacity": 5, "last updated": "01/11/2021", "ResponseCodes": ["xvar1", "xvar2", "xvar3", "yvar1", "yvar2", "yvar3" ] ,"distanceTOorigin": 530 }
My query which produces a normalised list of documents:
{
"query": {
"function_score": {
"query": { "match_all": {} },
"boost": "1",
"functions": [
{
"filter": [
{ "range": { "date":{ "gte": "01/11/2000" }}},
{ "terms_set": { "codes" : { "terms" : ["yvar1", "yvar2", "yvar3" ],
"minimum_should_match_script": { "source": "1" }}}}
],
"random_score": {},
"weight": 1
},
{
"filter": [
{ "terms_set": { "codes" : { "terms" : ["xvar1", "xvar2", "xvar3" ],
"minimum_should_match_script": { "source": "params.num_terms" }}}}
],
"weight": 1
},
{
"exp": {
"coordinates": {
"origin": "51.5132, -0.1362",
"offset": "0km",
"decay": 0.5,
"scale":"350km"}
},
"weight": 10
}
],
"max_boost": 10,
"score_mode": "max",
"boost_mode": "multiply"
}
}
}
This is what I tried as a query (substituting the match_all query) but does not work as I end up with a non-normalised list
"query": {
"bool": {
"should": [
{
"terms_set": { "codes" : { "terms" : ["xvar4"],
"minimum_should_match_script": { "source": "0" }, "boost" : 0.1}}
},
{
"match_all": {}
}
]
}
}
Any help for this ealsticsearch beginner will be greatly appreciated.

Answer 1

I found the solution by accessing the _score in a script_score query:
{
"query": {
"script_score": {
"query": {
"match": { "codes": "xvar4" }
},
"script": {
"source": "_score +0.1"
}
}
}
}