ivory.core.eval

Class GradedQrels

java.lang.Object

ivory.core.eval.Qrels

ivory.core.eval.GradedQrels

public class GradedQrels
extends Qrels

Representation of relevance judgments. In TREC parlance, qrels are judgments made by humans as to whether a document is relevant to an information need. Typically, qrels are created by a process known as "pooling" in large-scale system evaluations such as those at TREC.

Author:

Jimmy Lin, Lidan Wang

Constructor Summary

Constructors

Constructor and Description
GradedQrels(String file) Creates a GradedQrels object from a file

Method Summary

Methods

Modifier and Type	Method and Description
Set<String>	getReldocsForQid(String qid, boolean alsoReturnRating) Returns the set of relevant documents for a particular information need.
boolean	isRelevant(String qid, String docid) Determines if a document is relevant for a particular information need.

Methods inherited from class ivory.core.eval.Qrels

getQids, getReldocsForQid, helperHDFSTrainWSDTopics

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

GradedQrels

public GradedQrels(String file)

Creates a GradedQrels object from a file

Parameters:

file - file that contains the relevance judgments

Method Detail

getReldocsForQid

public Set<String> getReldocsForQid(String qid,
                           boolean alsoReturnRating)

Returns the set of relevant documents for a particular information need.

Parameters:

qid - the id of the information need

Returns:

the set of relevant documents

isRelevant

public boolean isRelevant(String qid,
                 String docid)

Determines if a document is relevant for a particular information need.

Overrides:

isRelevant in class Qrels

Parameters:

qid - id of the information need

docid - id of the document to test

Returns:

true if the document is relevant to the information need