Add Hyperlink-Induced Topic Search (HITS) as optional feature

Author: JohT

domains/anomaly-detection/anomalyDetectionCsv.sh

@@ -59,6 +59,9 @@ anomaly_detection_features() {
     # Determine the article rank if not already done
     execute_cypher_queries_until_results "${ANOMALY_DETECTION_FEATURE_CYPHER_DIR}/AnomalyDetectionFeature-ArticleRank-Exists.cypher" \
                                          "${ANOMALY_DETECTION_FEATURE_CYPHER_DIR}/AnomalyDetectionFeature-ArticleRank-Write.cypher" "${@}"
+    # Determine the HyperlinkInducedTopicSearch (HITS) Authority and Hub score if not already done
+    execute_cypher_queries_until_results "${ANOMALY_DETECTION_FEATURE_CYPHER_DIR}/AnomalyDetectionFeature-HyperlinkInducedTopicSearch-HITS-Exists.cypher" \
+                                         "${ANOMALY_DETECTION_FEATURE_CYPHER_DIR}/AnomalyDetectionFeature-HyperlinkInducedTopicSearch-HITS-Write.cypher" "${@}"
 }
 # Run queries to find anomalies in the graph.
 # 

domains/anomaly-detection/anomalyDetectionPython.sh

@@ -104,6 +104,9 @@ anomaly_detection_features() {
     # Determine the article rank if not already done
     execute_cypher_queries_until_results "${ANOMALY_DETECTION_FEATURE_CYPHER_DIR}/AnomalyDetectionFeature-ArticleRank-Exists.cypher" \
                                          "${ANOMALY_DETECTION_FEATURE_CYPHER_DIR}/AnomalyDetectionFeature-ArticleRank-Write.cypher" "${@}"
+    # Determine the HyperlinkInducedTopicSearch (HITS) Authority and Hub score if not already done
+    execute_cypher_queries_until_results "${ANOMALY_DETECTION_FEATURE_CYPHER_DIR}/AnomalyDetectionFeature-HyperlinkInducedTopicSearch-HITS-Exists.cypher" \
+                                         "${ANOMALY_DETECTION_FEATURE_CYPHER_DIR}/AnomalyDetectionFeature-HyperlinkInducedTopicSearch-HITS-Write.cypher" "${@}"
 }
 
 # Execute the Python scripts for anomaly detection.

domains/anomaly-detection/features/AnomalyDetectionFeature-HyperlinkInducedTopicSearch-HITS-Exists.cypher

@@ -0,0 +1,11 @@
+// Return the first node with a centralityHyperlinkInducedTopicSearch(Authority/Hub) (HITS) if it exists
+
+   MATCH (codeUnit)
+   WHERE $projection_node_label IN labels(codeUnit)
+     AND codeUnit.centralityHyperlinkInducedTopicSearchAuthority IS NOT NULL
+     AND codeUnit.centralityHyperlinkInducedTopicSearchHub       IS NOT NULL
+  RETURN codeUnit.name                       AS shortCodeUnitName
+        ,elementId(codeUnit)                 AS nodeElementId
+        ,codeUnit.centralityHyperlinkInducedTopicSearchAuthority AS hyperlinkInducedTopicSearchAuthority
+        ,codeUnit.centralityHyperlinkInducedTopicSearchHub       AS hyperlinkInducedTopicSearchHub
+   LIMIT 1

domains/anomaly-detection/features/AnomalyDetectionFeature-HyperlinkInducedTopicSearch-HITS-Write.cypher

@@ -0,0 +1,9 @@
+// Calculates and writes the Hyperlink-Induced Topic Search (HITS) centrality scores for anomaly detection
+
+  CALL gds.hits.write(
+ $projection_name + '-directed-cleaned', {
+     authProperty: 'centralityHyperlinkInducedTopicSearchAuthority'
+    ,hubProperty: 'centralityHyperlinkInducedTopicSearchHub'
+})
+YIELD nodePropertiesWritten, ranIterations, didConverge, preProcessingMillis, computeMillis, writeMillis
+RETURN nodePropertiesWritten, ranIterations, didConverge, preProcessingMillis, computeMillis, writeMillis