Elastic Search issue with simple_query_string - elasticsearch

Question

I could not find a perfect solution for the following situation, hope someone could help here.
Suppose there are following documents present for type called "groups":
{"name": "Orders 1.0"}
{"name": "Reports & Analysis 1.0"}
{"name": "Rebates 1.0"}
When I search documents using below simple_query_string I am getting all three records instead of only one records (i.e. #1)
{
"size" : 20,
"query" : {
"bool" : {
"must" : [
{
"bool" : {
"should" : [
{
"simple_query_string" : {
"query" : "Orders 1.0*",
"fields" : [
"name^1.0"
],
"flags" : -1,
"default_operator" : "or",
"lenient" : false,
"analyze_wildcard" : true,
"boost" : 1.0
}
}
],
"disable_coord" : false,
"adjust_pure_negative" : true,
"boost" : 1.0
}
}
],
"disable_coord" : false,
"adjust_pure_negative" : true,
"boost" : 1.0
}
}
}
I want only one record to be searched with name as Orders 1.0

Answer 1

The reason is this "default_operator" : "or",.
I assume that the field name is an analyzed field. If it uses an analyzer, then it will be tokenized. The same goes for your input. So you have the following (for the standard analyzer)
Orders 1.0 -> ["orders", "1.0"]
Reports & Analysis 1.0 -> ["reports", "analysis", "1.0"]
Rebates 1.0 -> ["rebates", "1.0"]
GET _analyze
{
"analyzer": "standard",
"text": ["Orders 1.0", "Reports & Analysis 1.0", "Rebates 1.0"]
}
{
"tokens" : [
{
"token" : "orders",
"start_offset" : 0,
"end_offset" : 6,
"type" : "<ALPHANUM>",
"position" : 0
},
{
"token" : "1.0",
"start_offset" : 7,
"end_offset" : 10,
"type" : "<NUM>",
"position" : 1
},
{
"token" : "reports",
"start_offset" : 11,
"end_offset" : 18,
"type" : "<ALPHANUM>",
"position" : 2
},
{
"token" : "analysis",
"start_offset" : 21,
"end_offset" : 29,
"type" : "<ALPHANUM>",
"position" : 3
},
{
"token" : "1.0",
"start_offset" : 30,
"end_offset" : 33,
"type" : "<NUM>",
"position" : 4
},
{
"token" : "rebates",
"start_offset" : 34,
"end_offset" : 41,
"type" : "<ALPHANUM>",
"position" : 5
},
{
"token" : "1.0",
"start_offset" : 42,
"end_offset" : 45,
"type" : "<NUM>",
"position" : 6
}
]
}
Since the default operator is OR, it means that you need any of the analyzed tokens to match, not all. And since you've included 1.0 that is present on all of them, it matches all the documents. One solution is to change the default operator to AND