/** Replace document refids with the actual references upon deserialization. */
private void documentIdToReference(Cluster cluster, Map<String, Document> documents) {
if (cluster.documentIds != null) {
for (Cluster.DocumentRefid documentRefid : cluster.documentIds) {
cluster.addDocuments(documents.get(documentRefid.refid));
}
}
for (Cluster subcluster : cluster.getSubclusters()) {
documentIdToReference(subcluster, documents);
}
}
private void addToCluster(Map<Object, Cluster> clusters, Object fieldValue, Document document) {
if (fieldValue == null) {
return;
}
Cluster cluster = clusters.get(fieldValue);
if (cluster == null) {
cluster = new Cluster();
cluster.addPhrases(buildClusterLabel(fieldValue));
clusters.put(fieldValue, cluster);
}
cluster.addDocuments(document);
}
/**
* Builds an "Other Topics" cluster that groups those documents from <code>allDocument</code> that
* were not referenced in any cluster in <code>clusters</code>.
*
* @param allDocuments all documents to check against
* @param clusters list of clusters with assigned documents
* @param label label for the "Other Topics" group
* @return the "Other Topics" cluster
*/
public static Cluster buildOtherTopics(
List<Document> allDocuments, List<Cluster> clusters, String label) {
final Set<Document> unclusteredDocuments = Sets.newLinkedHashSet(allDocuments);
final Set<Document> assignedDocuments = Sets.newHashSet();
for (Cluster cluster : clusters) {
collectAllDocuments(cluster, assignedDocuments);
}
unclusteredDocuments.removeAll(assignedDocuments);
final Cluster otherTopics = new Cluster(label);
otherTopics.addDocuments(unclusteredDocuments);
otherTopics.setOtherTopics(true);
return otherTopics;
}
