Skip to content

Instantly share code, notes, and snippets.

@brwe
Created October 29, 2013 10:43
Show Gist options
  • Save brwe/7212356 to your computer and use it in GitHub Desktop.
Save brwe/7212356 to your computer and use it in GitHub Desktop.
scoring oddities
#init data
DELETE testtags/
POST testtags/
{
"settings" : {
"index" : {
"number_of_shards" : 1,
"number_of_replicas" : 0
}
}
}
#oddities of text scoring
POST testtags/test/1
{
"location": ["berlin","munich","hamburg"]
}
POST testtags/test/3
{
"location": "berlin"
}
POST testtags/test/_search
{
"query": {
"match": {
"location": "berlin"
}
}
}
POST idfidx/test/2
{
"location": "berlin berlin"
}
POST idfidx/test/_search
{
"query": {
"match": {
"location": "berlin"
}
}
}
POST idfidx/test/_search
{
"query": {
"match": {
"location": "berlin kreuzberg"
}
}
}
POST idfidx/test/4
{
"location": "kreuzberg"
}
POST idfidx/test/_search
POST idfidx/test/_search
{
"query": {
"match": {
"location": "berlin kreuzberg"
}
}
}
# multi match - even more complications!
DELETE idfidx/
POST idfidx/
{
"settings" : {
"index" : {
"number_of_shards" : 1,
"number_of_replicas" : 0
}
}
}
POST idfidx/test/1
{
"name":"ice for responsible people that like pandas",
"food": "ice cream",
"location": "berlin kreuzberg"
}
POST idfidx/test/2
{
"name":"ice for a low price",
"food": "ice cream",
"location": "berlin kreuzberg (check kreuzberg out now while it is still cool!)"
}
POST idfidx/test/_search
{
"query": {
"multi_match": {
"query": "ice cream berlin kreuzberg",
"fields": ["food","location"]
}
}
}
#should be lower because of the field length, right? why both same score?
POST idfidx/test/1
{
"name":"out of nizza",
"food": "lobster oysters",
"location": "berlin kreuzberg"
}
POST idfidx/test/2
{
"name":"ice for a high price",
"food": "ice cream",
"location": "potsdam babelsberg"
}
POST idfidx/test/3
{
"name":"ice for a low price",
"food": "ice cream",
"location": "berlin kreuzberg"
}
POST idfidx/test/_search
POST idfidx/test/_search
{
"query": {
"multi_match": {
"query": "ice cream berlin kreuzberg",
"fields": ["food","location"]
}
},
"explain": false
}
# why ice cream in berlin kreuzberg last?
POST idfidx/test/4
{
"name":"karate",
"food": "sushi",
"location": "berlin kreuzberg"
}
DELETE idfidx/test/4
POST idfidx/test/_search
{
"query": {
"multi_match": {
"query": "ice cream berlin kreuzberg",
"fields": ["food","location"]
}
},
"explain": false
}
#why 1 and 4 lower? -> again idf! remember from before!
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment