svn commit: r1854991 - /jackrabbit/oak/trunk/oak-lucene/src/main/java/org/apache/jackrabbit/oak/plugins/index/lucene/util/fv/SimSearchUtils.java

[tommaso]

Author: tommaso
Date: Thu Mar  7 16:36:44 2019
New Revision: 1854991

URL: http://svn.apache.org/viewvc?rev=1854991&view=rev
Log:
OAK-8119 - use exact fv distance score for similarity search

Modified:
    
jackrabbit/oak/trunk/oak-lucene/src/main/java/org/apache/jackrabbit/oak/plugins/index/lucene/util/fv/SimSearchUtils.java

Modified: 
jackrabbit/oak/trunk/oak-lucene/src/main/java/org/apache/jackrabbit/oak/plugins/index/lucene/util/fv/SimSearchUtils.java
URL: 
http://svn.apache.org/viewvc/jackrabbit/oak/trunk/oak-lucene/src/main/java/org/apache/jackrabbit/oak/plugins/index/lucene/util/fv/SimSearchUtils.java?rev=1854991&r1=1854990&r2=1854991&view=diff
==============================================================================
--- 
jackrabbit/oak/trunk/oak-lucene/src/main/java/org/apache/jackrabbit/oak/plugins/index/lucene/util/fv/SimSearchUtils.java
 (original)
+++ 
jackrabbit/oak/trunk/oak-lucene/src/main/java/org/apache/jackrabbit/oak/plugins/index/lucene/util/fv/SimSearchUtils.java
 Thu Mar  7 16:36:44 2019
@@ -255,16 +255,19 @@ public class SimSearchUtils {
             if (binaryValue != null) {
                 double[] inputVector = toDoubleArray(binaryValue.bytes);
                 for (int j = 0; j < docs.scoreDocs.length; j++) {
-                    double[] currentVector = 
toDoubleArray(indexSearcher.doc(docs.scoreDocs[j].doc)
-                            .getBinaryValue(fieldName).bytes);
-                    double distance = dist(inputVector, currentVector) + 
1e-10; // constant term to avoid division by zero
-                    if (Double.isNaN(distance) || Double.isInfinite(distance)) 
{
-                        toDiscard.add(docs.scoreDocs[j].doc);
-                    } else {
-                        distSum += distance;
-                        counter++;
-                        distances.put(docs.scoreDocs[j].doc, distance);
-                        docs.scoreDocs[j].score += (float) (1d / distance); // 
additive similarity boosting
+                    BytesRef featureVectorBinary = 
indexSearcher.doc(docs.scoreDocs[j].doc)
+                            .getBinaryValue(fieldName);
+                    if (featureVectorBinary != null) {
+                        double[] currentVector = 
toDoubleArray(featureVectorBinary.bytes);
+                        double distance = dist(inputVector, currentVector) + 
1e-10; // constant term to avoid division by zero
+                        if (Double.isNaN(distance) || 
Double.isInfinite(distance)) {
+                            toDiscard.add(docs.scoreDocs[j].doc);
+                        } else {
+                            distSum += distance;
+                            counter++;
+                            distances.put(docs.scoreDocs[j].doc, distance);
+                            docs.scoreDocs[j].score = (float) (1d / distance);
+                        }
                     }
                 }
             }