honeymoose
/
OpenSearch
mirror of https://github.com/honeymoose/OpenSearch.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Normalize rated document key parameter 

To be consitent with the output of the search API, we should use the same field
names for specifying the document ("_index", "_type", "_id") when providing the
rated documents in the `rank_eval` request.
This commit is contained in:
Christoph Büscher 2016-09-21 12:36:50 +02:00
parent 306ea1c613
commit 6c1003b81a
4 changed files with 40 additions and 40 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
modules/rank-eval/src/main/java/org/elasticsearch/index/rankeval/RatedDocument.java
View File

@ -38,9 +38,9 @@ import java.util.Objects;
public class RatedDocument extends ToXContentToBytes implements Writeable {
public static final ParseField RATING_FIELD = new ParseField("rating");
public static final ParseField DOC_ID_FIELD = new ParseField("doc_id");
public static final ParseField TYPE_FIELD = new ParseField("type");
public static final ParseField INDEX_FIELD = new ParseField("index");
public static final ParseField DOC_ID_FIELD = new ParseField("_id");
public static final ParseField TYPE_FIELD = new ParseField("_type");
public static final ParseField INDEX_FIELD = new ParseField("_index");
private static final ConstructingObjectParser<RatedDocument, ParseFieldMatcherSupplier> PARSER =
new ConstructingObjectParser<>("rated_document",

6
modules/rank-eval/src/test/java/org/elasticsearch/index/rankeval/RatedRequestsTests.java
View File

@ -127,9 +127,9 @@ public class RatedRequestsTests extends ESTestCase {
+ " \"size\": 10\n"
+ " },\n"
+ " \"ratings\": [ "
+ " {\"index\": \"test\", \"type\": \"testtype\", \"doc_id\": \"1\", \"rating\" : 1 }, "
+ " {\"type\": \"testtype\", \"index\": \"test\", \"doc_id\": \"2\", \"rating\" : 0 }, "
+ " {\"doc_id\": \"3\", \"index\": \"test\", \"type\": \"testtype\", \"rating\" : 1 }]\n"
+ " {\"_index\": \"test\", \"_type\": \"testtype\", \"_id\": \"1\", \"rating\" : 1 }, "
+ " {\"_type\": \"testtype\", \"_index\": \"test\", \"_id\": \"2\", \"rating\" : 0 }, "
+ " {\"_id\": \"3\", \"_index\": \"test\", \"_type\": \"testtype\", \"rating\" : 1 }]\n"
+ "}";
XContentParser parser = XContentFactory.xContent(querySpecString).createParser(querySpecString);
QueryParseContext queryContext = new QueryParseContext(searchRequestParsers.queryParsers, parser, ParseFieldMatcher.STRICT);

20
modules/rank-eval/src/test/resources/rest-api-spec/test/rank_eval/10_basic.yaml
View File

@ -46,22 +46,22 @@
"id": "amsterdam_query",
"request": { "query": { "match" : {"text" : "amsterdam" }}},
"ratings": [
{"index": "foo", "type": "bar", "doc_id": "doc1", "rating": 0},
{"index": "foo", "type": "bar", "doc_id": "doc2", "rating": 1},
{"index": "foo", "type": "bar", "doc_id": "doc3", "rating": 1}]
{"_index": "foo", "_type": "bar", "_id": "doc1", "rating": 0},
{"_index": "foo", "_type": "bar", "_id": "doc2", "rating": 1},
{"_index": "foo", "_type": "bar", "_id": "doc3", "rating": 1}]
},
{
"id" : "berlin_query",
"request": { "query": { "match" : { "text" : "berlin" } }, "size" : 10 },
"ratings": [{"index": "foo", "type": "bar", "doc_id": "doc1", "rating": 1}]
"ratings": [{"_index": "foo", "_type": "bar", "_id": "doc1", "rating": 1}]
}
],
"metric" : { "precisionatn": { "size": 10}}
}
- match: {rank_eval.quality_level: 1}
- match: {rank_eval.unknown_docs.amsterdam_query: [ {"index": "foo", "type": "bar", "doc_id": "doc4"}]}
- match: {rank_eval.unknown_docs.berlin_query: [ {"index": "foo", "type": "bar", "doc_id": "doc4"}]}
- match: {rank_eval.unknown_docs.amsterdam_query: [ {"_index": "foo", "_type": "bar", "_id": "doc4"}]}
- match: {rank_eval.unknown_docs.berlin_query: [ {"_index": "foo", "_type": "bar", "_id": "doc4"}]}
---
"Reciprocal Rank":
@ -111,13 +111,13 @@
"id": "amsterdam_query",
"request": { "query": { "match" : {"text" : "amsterdam" }}},
# doc4 should be returned in third position, so reciprocal rank is 1/3
"ratings": [{"index": "foo", "type": "bar", "doc_id": "doc4", "rating": 1}]
"ratings": [{"_index": "foo", "_type": "bar", "_id": "doc4", "rating": 1}]
},
{
"id" : "berlin_query",
"request": { "query": { "match" : { "text" : "berlin" } }, "size" : 10 },
# doc1 should be returned in first position, doc3 in second, so reciprocal rank is 1/2
"ratings": [{"index": "foo", "type": "bar", "doc_id": "doc4", "rating": 1}]
"ratings": [{"_index": "foo", "_type": "bar", "_id": "doc4", "rating": 1}]
}
],
"metric" : { "reciprocal_rank": {} }
@ -134,13 +134,13 @@
"id": "amsterdam_query",
"request": { "query": { "match" : {"text" : "amsterdam" }}},
# doc4 should be returned in third position, so reciprocal rank is 1/3
"ratings": [{"index": "foo", "type": "bar", "doc_id": "doc4", "rating": 1}]
"ratings": [{"_index": "foo", "_type": "bar", "_id": "doc4", "rating": 1}]
},
{
"id" : "berlin_query",
"request": { "query": { "match" : { "text" : "berlin" } }, "size" : 10 },
# doc1 should be returned in first position, doc3 in second, so reciprocal rank is 1/2
"ratings": [{"index": "foo", "type": "bar", "doc_id": "doc4", "rating": 1}]
"ratings": [{"_index": "foo", "_type": "bar", "_id": "doc4", "rating": 1}]
}
],
"metric" : {

48
modules/rank-eval/src/test/resources/rest-api-spec/test/rank_eval/20_dcg.yaml
View File

@ -52,12 +52,12 @@
"id": "dcg_query",
"request": { "query": { "match_all" : {}}, "sort" : [ "bar" ] },
"ratings": [
{"index": "foo", "type": "bar", "doc_id": "doc1", "rating": 3},
{"index": "foo", "type": "bar", "doc_id": "doc2", "rating": 2},
{"index": "foo", "type": "bar", "doc_id": "doc3", "rating": 3},
{"index": "foo", "type": "bar", "doc_id": "doc4", "rating": 0},
{"index": "foo", "type": "bar", "doc_id": "doc5", "rating": 1},
{"index": "foo", "type": "bar", "doc_id": "doc6", "rating": 2}]
{"_index" : "foo", "_type" : "bar", "_id" : "doc1", "rating": 3},
{"_index" : "foo", "_type" : "bar", "_id" : "doc2", "rating": 2},
{"_index" : "foo", "_type" : "bar", "_id" : "doc3", "rating": 3},
{"_index" : "foo", "_type" : "bar", "_id" : "doc4", "rating": 0},
{"_index" : "foo", "_type" : "bar", "_id" : "doc5", "rating": 1},
{"_index" : "foo", "_type" : "bar", "_id" : "doc6", "rating": 2}]
}
],
"metric" : { "dcg_at_n": { "size": 6}}
@ -75,12 +75,12 @@
"id": "dcg_query_reverse",
"request": { "query": { "match_all" : {}}, "sort" : [ {"bar" : "desc" }] },
"ratings": [
{"index": "foo", "type": "bar", "doc_id": "doc1", "rating": 3},
{"index": "foo", "type": "bar", "doc_id": "doc2", "rating": 2},
{"index": "foo", "type": "bar", "doc_id": "doc3", "rating": 3},
{"index": "foo", "type": "bar", "doc_id": "doc4", "rating": 0},
{"index": "foo", "type": "bar", "doc_id": "doc5", "rating": 1},
{"index": "foo", "type": "bar", "doc_id": "doc6", "rating": 2}]
{"_index" : "foo", "_type" : "bar", "_id" : "doc1", "rating": 3},
{"_index" : "foo", "_type" : "bar", "_id" : "doc2", "rating": 2},
{"_index" : "foo", "_type" : "bar", "_id" : "doc3", "rating": 3},
{"_index" : "foo", "_type" : "bar", "_id" : "doc4", "rating": 0},
{"_index" : "foo", "_type" : "bar", "_id" : "doc5", "rating": 1},
{"_index" : "foo", "_type" : "bar", "_id" : "doc6", "rating": 2}]
},
],
"metric" : { "dcg_at_n": { "size": 6}}
@ -98,23 +98,23 @@
"id": "dcg_query",
"request": { "query": { "match_all" : {}}, "sort" : [ "bar" ] },
"ratings": [
{"index": "foo", "type": "bar", "doc_id": "doc1", "rating": 3},
{"index": "foo", "type": "bar", "doc_id": "doc2", "rating": 2},
{"index": "foo", "type": "bar", "doc_id": "doc3", "rating": 3},
{"index": "foo", "type": "bar", "doc_id": "doc4", "rating": 0},
{"index": "foo", "type": "bar", "doc_id": "doc5", "rating": 1},
{"index": "foo", "type": "bar", "doc_id": "doc6", "rating": 2}]
{"_index" : "foo", "_type" : "bar", "_id" : "doc1", "rating": 3},
{"_index" : "foo", "_type" : "bar", "_id" : "doc2", "rating": 2},
{"_index" : "foo", "_type" : "bar", "_id" : "doc3", "rating": 3},
{"_index" : "foo", "_type" : "bar", "_id" : "doc4", "rating": 0},
{"_index" : "foo", "_type" : "bar", "_id" : "doc5", "rating": 1},
{"_index" : "foo", "_type" : "bar", "_id" : "doc6", "rating": 2}]
},
{
"id": "dcg_query_reverse",
"request": { "query": { "match_all" : {}}, "sort" : [ {"bar" : "desc" }] },
"ratings": [
{"index": "foo", "type": "bar", "doc_id": "doc1", "rating": 3},
{"index": "foo", "type": "bar", "doc_id": "doc2", "rating": 2},
{"index": "foo", "type": "bar", "doc_id": "doc3", "rating": 3},
{"index": "foo", "type": "bar", "doc_id": "doc4", "rating": 0},
{"index": "foo", "type": "bar", "doc_id": "doc5", "rating": 1},
{"index": "foo", "type": "bar", "doc_id": "doc6", "rating": 2}]
{"_index" : "foo", "_type" : "bar", "_id" : "doc1", "rating": 3},
{"_index" : "foo", "_type" : "bar", "_id" : "doc2", "rating": 2},
{"_index" : "foo", "_type" : "bar", "_id" : "doc3", "rating": 3},
{"_index" : "foo", "_type" : "bar", "_id" : "doc4", "rating": 0},
{"_index" : "foo", "_type" : "bar", "_id" : "doc5", "rating": 1},
{"_index" : "foo", "_type" : "bar", "_id" : "doc6", "rating": 2}]
},
],
"metric" : { "dcg_at_n": { "size": 6}}