Synonym search in ElasticSearch - elasticsearch

Question

I want to retrieve the data from the index using the notion of synonym. When I perform a search with title A I also want to retrieve the documents whose title contains B. For that I set up the following mapping :
{
"settings": {
"index" : {
"analysis" : {
"filter" : {
"synonym_filter" : {
"type" : "synonym",
"synonyms" : [
"A=>A,B"
]
}
},
"analyzer" : {
"synonym_analyzer" : {
"tokenizer" : "keyword",
"filter" : ["synonym_filter"]
}
}
}
}
},
"mappings": {
"properties": {
"title": {
"type": "text",
"analyzer" : "synonym_analyzer"
}
}
}
}
I then added 3 documents to my index
{
"title": "C"
}
{
"title": "B"
}
{
"title": "A"
}
I then used the analysis api to see if it works (everything is ok):
curl -X GET "localhost:9200/my_custom_index_title/_analyze?pretty" -H 'Content-Type: application/json' -d'
{
"analyzer": "synonym_analyzer",
"text": "A"
}
'
{
"tokens" : [
{
"token" : "A",
"start_offset" : 0,
"end_offset" : 1,
"type" : "SYNONYM",
"position" : 0
},
{
"token" : "B",
"start_offset" : 0,
"end_offset" : 1,
"type" : "SYNONYM",
"position" : 0
}
]
}
url -X GET "localhost:9200/my_custom_index_title/_analyze?pretty" -H 'Content-Type: application/json' -d'
{
"analyzer": "synonym_analyzer",
"text": "B"
}
'
{
"tokens" : [
{
"token" : "B",
"start_offset" : 0,
"end_offset" : 1,
"type" : "word",
"position" : 0
}
]
}
When I search for title A results are correct :
{
"query": {
"match": {
"title": {
"query": "A"
}
}
}
}
{
"took": 1,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 2,
"relation": "eq"
},
"max_score": 0.6951314,
"hits": [
{
"_index": "my_custom_index_title",
"_id": "i5bb_4IBqFAXxSLAgrDj",
"_score": 0.6951314,
"_source": {
"title": "A"
}
},
{
"_index": "my_custom_index_title",
"_id": "jJbb_4IBqFAXxSLAlLBj",
"_score": 0.52354836,
"_source": {
"title": "B"
}
}
]
}
}
But when I search for B the results are not correct, I just want result who contains B when I search and not A
{
"query": {
"match": {
"title": {
"query": "B"
}
}
}
}
{
"took": 1,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 2,
"relation": "eq"
},
"max_score": 0.52354836,
"hits": [
{
"_index": "my_custom_index_title",
"_id": "i5bb_4IBqFAXxSLAgrDj",
"_score": 0.52354836,
"_source": {
"title": "A"
}
},
{
"_index": "my_custom_index_title",
"_id": "jJbb_4IBqFAXxSLAlLBj",
"_score": 0.52354836,
"_source": {
"title": "B"
}
}
]
}
}
For example when I search for computer I wish to obtain laptop, computer, mac. But when I search for mac I only want to get the results for it (not laptop and computer)
I do not understand why the result for the search with B does not return only one result

Answer 1

I understand, in this case as you applied synonym_analyzer as a field analyzer, you indexed the synonyms.
To solve it, you can use synonyms only at search time, adding the parameter "search_analyzer". Note that I added the lowercase filter in the synonym_analyzer because the standard analyzer applies lowercase by default.
To get token synonyms for Term B do this:
{
"settings": {
"index": {
"analysis": {
"filter": {
"synonym_filter": {
"type": "synonym",
"expand":"false",
"synonyms": [
"A=>A,B"
]
}
},
"analyzer": {
"synonym_analyzer": {
"tokenizer": "keyword",
"filter": [
"lowercase",
"synonym_filter"
]
}
}
}
}
},
"mappings": {
"properties": {
"title": {
"type": "text",
"analyzer": "standard",
"search_analyzer": "synonym_analyzer"
}
}
}
}