Package org.elasticsearch.index.rankeval
Class RecallAtK
java.lang.Object
org.elasticsearch.index.rankeval.RecallAtK
- All Implemented Interfaces:
org.elasticsearch.common.io.stream.NamedWriteable,org.elasticsearch.common.io.stream.Writeable,EvaluationMetric,org.elasticsearch.xcontent.ToXContent,org.elasticsearch.xcontent.ToXContentObject
Metric implementing Recall@K
(https://en.wikipedia.org/wiki/Evaluation_measures_(information_retrieval)#Recall).
By default documents with a rating equal or bigger than 1 are considered to be "relevant" for this calculation. This value can be changed using the `relevant_rating_threshold` parameter.
The `k` parameter (defaults to 10) controls the search window size.
By default documents with a rating equal or bigger than 1 are considered to be "relevant" for this calculation. This value can be changed using the `relevant_rating_threshold` parameter.
The `k` parameter (defaults to 10) controls the search window size.
-
Nested Class Summary
Nested ClassesNested classes/interfaces inherited from interface org.elasticsearch.xcontent.ToXContent
org.elasticsearch.xcontent.ToXContent.DelegatingMapParams, org.elasticsearch.xcontent.ToXContent.MapParams, org.elasticsearch.xcontent.ToXContent.Params -
Field Summary
FieldsFields inherited from interface org.elasticsearch.xcontent.ToXContent
EMPTY_PARAMS -
Constructor Summary
Constructors -
Method Summary
Modifier and TypeMethodDescriptionfinal booleanevaluate(String taskId, org.elasticsearch.search.SearchHit[] hits, List<RatedDocument> ratedDocs) Compute recall at k based on provided relevant document IDs.Metrics can define a size of the search hits windows they want to retrieve by overwriting this method.static RecallAtKfromXContent(org.elasticsearch.xcontent.XContentParser parser) intgetK()intReturn the rating threshold above which ratings are considered to be "relevant" for this metric.final inthashCode()org.elasticsearch.xcontent.XContentBuildertoXContent(org.elasticsearch.xcontent.XContentBuilder builder, org.elasticsearch.xcontent.ToXContent.Params params) voidwriteTo(org.elasticsearch.common.io.stream.StreamOutput out) Methods inherited from class java.lang.Object
clone, finalize, getClass, notify, notifyAll, toString, wait, wait, waitMethods inherited from interface org.elasticsearch.index.rankeval.EvaluationMetric
combineMethods inherited from interface org.elasticsearch.xcontent.ToXContentObject
isFragment
-
Field Details
-
NAME
- See Also:
-
-
Constructor Details
-
RecallAtK
public RecallAtK(int relevantRatingThreshold, int k) Metric implementing Recall@K.- Parameters:
relevantRatingThreshold- ratings equal or above this value will be considered relevant.k- controls the window size for the search results the metric takes into account
-
RecallAtK
public RecallAtK()
-
-
Method Details
-
fromXContent
-
writeTo
- Specified by:
writeToin interfaceorg.elasticsearch.common.io.stream.Writeable- Throws:
IOException
-
toXContent
public org.elasticsearch.xcontent.XContentBuilder toXContent(org.elasticsearch.xcontent.XContentBuilder builder, org.elasticsearch.xcontent.ToXContent.Params params) throws IOException - Specified by:
toXContentin interfaceorg.elasticsearch.xcontent.ToXContent- Throws:
IOException
-
getWriteableName
- Specified by:
getWriteableNamein interfaceorg.elasticsearch.common.io.stream.NamedWriteable
-
getRelevantRatingThreshold
public int getRelevantRatingThreshold()Return the rating threshold above which ratings are considered to be "relevant" for this metric. Defaults to 1. -
getK
public int getK() -
forcedSearchSize
Description copied from interface:EvaluationMetricMetrics can define a size of the search hits windows they want to retrieve by overwriting this method. The default implementation returns an empty optional.- Specified by:
forcedSearchSizein interfaceEvaluationMetric- Returns:
- the number of search hits this metrics requests
-
evaluate
public EvalQueryQuality evaluate(String taskId, org.elasticsearch.search.SearchHit[] hits, List<RatedDocument> ratedDocs) Compute recall at k based on provided relevant document IDs.- Specified by:
evaluatein interfaceEvaluationMetric- Parameters:
taskId- an identifier of the query for which the search ranking is evaluatedhits- the search result hitsratedDocs- the documents that contain the document rating for this query case- Returns:
- recall at k for above
SearchResultlist.
-
equals
-
hashCode
public final int hashCode()
-