com.graphaware.nlp.ml.word2vec.Word2VecIndexCreator Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of nlp Show documentation
GraphAware Framework Module for adding NLP capabilities to Neo4j.
There is a newer version: 3.5.4.53.18
/*
 * Copyright (c) 2013-2018 GraphAware
 *
 * This file is part of the GraphAware Framework.
 *
 * GraphAware Framework is free software: you can redistribute it and/or modify it under the terms of
 * the GNU General Public License as published by the Free Software Foundation, either
 * version 3 of the License, or (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;
 * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 * See the GNU General Public License for more details. You should have received a copy of
 * the GNU General Public License along with this program.  If not, see
 * .
 */

package com.graphaware.nlp.ml.word2vec;

import com.graphaware.common.log.LoggerFactory;
import com.graphaware.nlp.util.TypeConverter;
import org.apache.commons.io.FileUtils;
import org.apache.commons.io.LineIterator;
import org.apache.lucene.analysis.Analyzer;
import org.apache.lucene.analysis.core.KeywordAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.document.StoredField;
import org.apache.lucene.document.StringField;
import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.index.IndexWriterConfig.OpenMode;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.FSDirectory;
import org.neo4j.logging.Log;

import java.io.File;
import java.io.IOException;
import java.nio.file.Paths;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;

public class Word2VecIndexCreator {
    private static final Log LOG = LoggerFactory.getLogger(Word2VecIndexLookup.class);

    public static final String VECTOR_FIELD = "values";
    public static final String WORD_FIELD = "word";
    public static final String NEAREST_NEIGHBORS_FIELD = "nn";

    public static boolean loadFromFile(String sourceFile, String indexPath, boolean create, String language) {
        try {
            if (!create) {
                File f = new File(indexPath);
                if (f.exists()) {
                    return true;
                }
            }
            Directory dir = FSDirectory.open(Paths.get(indexPath));
            Analyzer analyzer = new KeywordAnalyzer();
            IndexWriterConfig iwc = new IndexWriterConfig(analyzer);
            if (create) {
                iwc.setOpenMode(OpenMode.CREATE);
            } else {
                iwc.setOpenMode(OpenMode.CREATE_OR_APPEND);
            }
            iwc.setRAMBufferSizeMB(256.0);
            try (IndexWriter writer = new IndexWriter(dir, iwc)) {
                LOG.info("Starting indexing in " + indexPath);
                indexWord2Vec(writer, sourceFile, language);
                writer.forceMerge(1);
            }
            
        }   catch (IOException ex) {
            LOG.error("Error loading Word2Vec: " + sourceFile, ex);
            return false;
        }
        return true;
    }

    private static void indexWord2Vec(IndexWriter writer, String sourceFile, String language) throws IOException {

        LineIterator it = FileUtils.lineIterator(new File(sourceFile), "UTF-8");
        try {
            while (it.hasNext()) {
                String line = it.nextLine();
                String[] split = line.split(" ");
                if (split != null && split.length > 2) {
                    Document doc = new Document();
                    String word = split[0];
                    String wordToUse = split[0];
                    if (word.startsWith("/c/") && !word.startsWith("/c/" + language + "/")) {
                        continue;
                    }
                    if (word.startsWith("/c/" + language)) {
                        wordToUse = wordToUse.replace("/c/" + language + "/", "").trim();
                    }
                    if (!wordToUse.matches("[\\p{L}]+")) {
                        continue;
                    }
                    doc.add(new StringField(WORD_FIELD, wordToUse, Field.Store.YES));
                    float[] vector = new float[split.length - 1];
                    for (int i = 0; i < split.length - 1; i++) {
                        vector[i] = Float.parseFloat(split[i + 1]);
                    }
                    doc.add(new StoredField(VECTOR_FIELD, TypeConverter.toByteArray(vector)));
                    writer.addDocument(doc);
                }
            }
        } finally {
            it.close();
        }
    }
    
    public static List inspectDirectoryAndLoad(String path, String destPath, String language) {
        LOG.info("Inspect directories and LOAD");

        List modelNames = new ArrayList<>();
        if (path == null || path.length() == 0) {
            LOG.error("Scanning for word2Vec files: wrong path specified.");
            return modelNames;
        }
        File folder = new File(path);
        File[] listOfFiles = folder.listFiles();
        if (listOfFiles == null) {
            LOG.warn("No files in " + path + " for loading word2Vec");
            return modelNames;
        }
        LOG.info("path = " + path);
        if (!path.endsWith("/")) {
            path = path + "/";
        }
        
        for (File file : listOfFiles) {
            if (!file.isFile()) {
                continue;
            }
            String fileName = file.getName();
            if (isIgnorableFile(fileName)) {
                continue;
            }
            String[] sp = fileName.split("-");
            String modelName = sp[0];
            LOG.info("Custom models: Found file " + fileName + ". Assigned name: " + modelName);
            if (loadFromFile((path + fileName), (destPath +  modelName), false, language)) {
                modelNames.add(modelName);
            }
        }

        return modelNames;
    }

    private static boolean isIgnorableFile(String filename) {
        List ignores = Arrays.asList(".DS_Store");
        if (!filename.contains(".txt")) {
            return true;
        } else {
            return ignores.contains(filename);
        }
    }
}