Term-document matrix in Lucene

Question

I am trying to get a term-document matrix from Lucene. It seems that most of the SO questions are for outdated APIs with different classes. I tried combining insight from these two questions to get a term vector from every document:

Relevant code, but DocEnum is not recognized in the current API. How can I get a term vector or count of all terms for every document?

IndexReader reader = DirectoryReader.open(index);

for (int i = 0;  i < reader.maxDoc(); i++) {
    Document doc = reader.document(i);
    Terms terms = reader.getTermVector(i, "country_text");

    if (terms != null && terms.size() > 0) {
        // access the terms for this field
        TermsEnum termsEnum = terms.iterator(); 
        BytesRef term = null;

        // explore the terms for this field
        while ((term = termsEnum.next()) != null) {
            // enumerate through documents, in this case only one
            DocsEnum docsEnum = termsEnum.docs(null, null); 
            int docIdEnum;
            while ((docIdEnum = docsEnum.nextDoc()) != DocIdSetIterator.NO_MORE_DOCS) {
                // get the term frequency in the document 
                System.out.println(term.utf8ToString()+ " " + docIdEnum + " " + docsEnum.freq()); 
            }
        }
    }
}

Full code:

import java.io.*;
import java.util.Iterator;

import org.apache.lucene.analysis.standard.StandardAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.document.StringField;
import org.apache.lucene.document.TextField;
import org.apache.lucene.index.DirectoryReader;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.index.Term;
import org.apache.lucene.index.Terms;
import org.apache.lucene.index.TermsEnum;
import org.apache.lucene.queryparser.classic.ParseException;
import org.apache.lucene.queryparser.classic.QueryParser;
import org.apache.lucene.search.BooleanClause;
import org.apache.lucene.search.BooleanQuery;
import org.apache.lucene.search.FuzzyQuery;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.ScoreDoc;
import org.apache.lucene.search.TermQuery;
import org.apache.lucene.search.TopDocs;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.RAMDirectory;
import org.apache.lucene.util.BytesRef;
import org.json.simple.JSONArray;
import org.json.simple.JSONObject;
import org.json.simple.JSONValue;
import org.json.simple.parser.JSONParser;

public class LuceneIndex {

    public static void main(String[] args) throws IOException, ParseException {

        String jsonFilePath = "wiki_data.json";
        JSONParser parser = new JSONParser();
        // Specify the analyzer for tokenizing text.
        StandardAnalyzer analyzer = new StandardAnalyzer();
        // create the index
        Directory index = new RAMDirectory();
        IndexWriterConfig config = new IndexWriterConfig(analyzer);
        IndexWriter w = new IndexWriter(index, config);

        try {     
            JSONArray a = (JSONArray) parser.parse(new FileReader(jsonFilePath));

            for (Object o : a) {
                JSONObject country = (JSONObject) o;
                String countryName = (String) country.get("country_name");
                String cityName = (String) country.get("city_name");
                String countryText = (String) country.get("country_text");
                String cityText = (String) country.get("city_text");
                System.out.println(cityName);
                addDoc(w, countryName, cityName, countryText, cityText);
            }
            w.close();

            IndexReader reader = DirectoryReader.open(index);

            for (int i = 0;  i < reader.maxDoc(); i++) {
                Document doc = reader.document(i);
                Terms terms = reader.getTermVector(i, "country_text");

                if (terms != null && terms.size() > 0) {
                    // access the terms for this field
                    TermsEnum termsEnum = terms.iterator(); 
                    BytesRef term = null;

                    // explore the terms for this field
                    while ((term = termsEnum.next()) != null) {
                        // enumerate through documents, in this case only one
                        DocsEnum docsEnum = termsEnum.docs(null, null); 
                        int docIdEnum;
                        while ((docIdEnum = docsEnum.nextDoc()) != DocIdSetIterator.NO_MORE_DOCS) {
                            // get the term frequency in the document 
                            System.out.println(term.utf8ToString()+ " " + docIdEnum + " " + docsEnum.freq()); 
                        }
                    }
                }
            }

            // reader can be closed when there
            // is no need to access the documents any more.
            reader.close();

        } catch (FileNotFoundException e) {
            e.printStackTrace();
        } catch (IOException e) {
            e.printStackTrace();
        } catch (org.json.simple.parser.ParseException e) {
            e.printStackTrace();
        }
    }

    private static void addDoc(IndexWriter w, String countryName, String cityName, 
            String countryText, String cityText) throws IOException {
        Document doc = new Document();
        doc.add(new StringField("country_name", countryName, Field.Store.YES));
        doc.add(new StringField("city_name", cityName, Field.Store.YES));
        doc.add(new TextField("country_text", countryText, Field.Store.YES));
        doc.add(new TextField("city_text", cityText, Field.Store.YES));

        w.addDocument(doc);
    }

}

score 1 · Answer 1 · answered May 08 '18 at 09:18

First thank for your code I had a little bug and your code helped me to complete it.

For me it works with this: (Lucene 7.2.1)

for(int i = 0; i < reader.maxDoc(); i++){
    Document doc = reader.document(i);
    Terms terms = reader.getTermVector(i, "text");

    if (terms != null && terms.size() > 0) {
        // access the terms for this field
        TermsEnum termsEnum = terms.iterator();
        BytesRef term = null;

        // explore the terms for this field
        while ((term = termsEnum.next()) != null) {
            // enumerate through documents, in this case only one
            PostingsEnum docsEnum = termsEnum.postings(null); 
            int docIdEnum;
            while ((docIdEnum = docsEnum.nextDoc()) != DocIdSetIterator.NO_MORE_DOCS) {
                // get the term frequency in the document
                System.out.println(term.utf8ToString()+ " " + docIdEnum + " " + docsEnum.freq());
            }
        }
    }
}

The Change here is I used PostingsEnum. DocsEnum is not available in Lucene 7.2.1 anymore.

But why it didn't work for you is how you add your document:

private void addDoc(IndexWriter w, String text, String name, String id) throws IOException {
    Document doc = new Document();
    // Create own FieldType to store Term Vectors
    FieldType ft = new FieldType();
    ft.setIndexOptions(IndexOptions.DOCS_AND_FREQS);
    ft.setTokenized(true);
    ft.setStored(true);
    ft.setStoreTermVectors(true);  //Store Term Vectors
    ft.freeze();
    StoredField t = new StoredField("text",text,ft);
    doc.add(t);


    doc.add(new StringField("name", name, Field.Store.YES));
    doc.add(new StringField("id", id, Field.Store.YES));
    w.addDocument(doc);
}

You have to create your own FieldType. None of the standard ones will save the term vectors.

score 0 · Answer 2 · answered Nov 23 '17 at 05:08

0

According to this question you should not use TextField for term frequency. Because it does not calculate it. Use "Field".

answered Nov 23 '17 at 05:08

Orçun Çolak

182
1
1
7

score 0 · Answer 3 · answered Jun 17 '20 at 23:31

0

You can also do it by make your field as this:

 FieldType myFieldType = new FieldType(TextField.TYPE_STORED);
 myFieldType.setStoreTermVectors(true);

Then re-index your documents. finally can get term vector!

answered Jun 17 '20 at 23:31

Hamed Sanaei

121
13

Term-document matrix in Lucene

3 Answers3

Linked