Example usage for org.apache.lucene.search.spell JaroWinklerDistance JaroWinklerDistance

Introduction

In this page you can find the example usage for org.apache.lucene.search.spell JaroWinklerDistance JaroWinklerDistance.

Prototype

public JaroWinklerDistance() 

Source Link

Document

Creates a new distance metric with the default threshold for the Jaro Winkler bonus (0.7)

Usage

From source file:com.mythesis.userbehaviouranalysis.ProfileAnalysis.java

License:Apache License

/**
 * a method that stores the query that has been suggested by the user
 * @param crawlerOutputPath SWebRank output directory used to check if a relevant query already exists
 * @param profile the query's relevant profile
 * @param query the given query//from  w  ww . j a  v a2s. c  om
 */
public void storeQuery(String crawlerOutputPath, String profile, String query) {

    System.out.println(crawlerOutputPath);
    System.out.println(profile);
    System.out.println(query);
    //Find output paths
    File root = new File(crawlerOutputPath);
    File[] contents = root.listFiles();
    List<String> sWebRanklevels = new ArrayList<>();
    for (File f : contents) {
        if (f.getAbsolutePath().contains("level"))
            sWebRanklevels.add(f.getAbsolutePath());
    }

    //Find all query paths
    List<String> queries = new ArrayList<>();
    for (String s : sWebRanklevels) {
        File level = new File(s);
        File[] queriesFiles = level.listFiles();
        for (File f : queriesFiles) {
            if (!f.getAbsolutePath().contains("txt")) {
                String str = f.getAbsolutePath();
                queries.add(str.substring(str.lastIndexOf("\\") + 1).replace("-query", "").replace("+", " "));
            }
        }
    }

    //check if a relevant query already exists - I use Jaro-Winkler distance
    query = query.trim().replaceAll(" +", " ");
    for (String q : queries) {
        JaroWinklerDistance jwd = new JaroWinklerDistance();
        double distance = jwd.getDistance(q, query);
        if (distance > 0.9) { // threshold = 0.9
            return;
        }
    }

    Mongo mongo = new Mongo("localhost", 27017);
    DB db = mongo.getDB("profileAnalysis");

    DBCollection DBqueries = db.getCollection("newQueries");
    BasicDBObject searchQuery = new BasicDBObject();
    searchQuery.put("profile", profile);
    DBObject document = DBqueries.findOne(searchQuery);
    boolean flag = false;

    //check if a relevant query exists in the database - I use Jaro-Winkler distance
    if (document != null) {
        flag = true;
        BasicDBList storedQueries = (BasicDBList) document.get("queries");
        for (Object quer : storedQueries) {
            JaroWinklerDistance jwd = new JaroWinklerDistance();
            double distance = jwd.getDistance((String) quer, query);
            if (distance > 0.9) { // threshold = 0.9
                return;
            }
        }
    }

    //if document already exists add the new query
    if (flag) {
        DBqueries.update(searchQuery, new BasicDBObject("$push", new BasicDBObject("queries", query)));
    } else { //otherwise create a new document
        BasicDBList dbl = new BasicDBList();
        dbl.add(query);
        BasicDBObject entry = new BasicDBObject("profile", profile).append("queries", dbl);
        DBqueries.insert(entry);
    }
}

---

From source file:fastcampus.lucene.example.search.SpellCheckerExample.java

License:Apache License

public static void main(String[] args) throws Exception {
    Directory directory = FSDirectory.open(Paths.get("./index/spell/"));
    SpellChecker spellChecker = new SpellChecker(directory);

    //Analyzer analyzer = new StandardAnalyzer();                             // ? 
    Analyzer analyzer = new Analyzer() {
        @Override//from w w  w.ja  v  a 2 s . c o  m
        protected TokenStreamComponents createComponents(String s) {
            Reader reader = new StringReader(s);
            Tokenizer tokenizer = new StandardTokenizer();
            tokenizer.setReader(reader);
            String name = "nfc_cf";
            Normalizer2 normalizer = Normalizer2.getInstance(null, name, Normalizer2.Mode.DECOMPOSE);
            TokenFilter filter = new ICUNormalizer2Filter(tokenizer, normalizer);
            return new TokenStreamComponents(tokenizer, filter);
        }
    };

    IndexWriterConfig indexWriterConfig = new IndexWriterConfig(analyzer); //?? Writer? ?  ?

    Path path = Paths.get("./data/spell/dic.txt");

    spellChecker.setSpellIndex(directory);
    spellChecker.clearIndex();
    spellChecker.indexDictionary(new PlainTextDictionary(path), indexWriterConfig, true);
    String wordForSuggestions = "?";
    //spellChecker.setStringDistance(new LevensteinDistance());  //#Levenstein  
    spellChecker.setStringDistance(new JaroWinklerDistance()); //Jaro-Winkler 

    int suggestionsNumber = 1;
    String[] suggestions = spellChecker.suggestSimilar(wordForSuggestions, suggestionsNumber);
    if (suggestions != null && suggestions.length > 0) {

        for (String word : suggestions) {

            System.out.println("Did you mean:" + word);

        }

    } else {

        System.out.println("No suggestions found for word:" + wordForSuggestions);

    }

}

---

From source file:lu.list.itis.dkd.aig.SimilarityProvider.java

License:Apache License

/**
 * Method used to compute a string similarity using Jaro Winkler as implemented by Apache's
 * Lucene./*w ww  .  java2 s.  c o m*/
 *
 * @param _this
 *        The first variable to compute similarity with.
 * @param that
 *        The second variable to compute similarity with.
 * @return The string similarity of both variables.
 * @see JaroWinklerDistance#getDistance(String, String)
 */
private float computeStringSimilarity(final Variable _this, final Variable that) {
    if (!useStringSimilarity) {
        return 0f;
    }

    return new JaroWinklerDistance().getDistance(_this.getTextContent(), that.getTextContent());
}

---

From source file:org.elasticsearch.search.suggest.phrase.DirectCandidateGeneratorBuilder.java

License:Apache License

private static StringDistance resolveDistance(String distanceVal) {
    distanceVal = distanceVal.toLowerCase(Locale.US);
    if ("internal".equals(distanceVal)) {
        return DirectSpellChecker.INTERNAL_LEVENSHTEIN;
    } else if ("damerau_levenshtein".equals(distanceVal) || "damerauLevenshtein".equals(distanceVal)) {
        return new LuceneLevenshteinDistance();
    } else if ("levenstein".equals(distanceVal)) {
        return new LevensteinDistance();
        // TODO Jaro and Winkler are 2 people - so apply same naming logic
        // as damerau_levenshtein
    } else if ("jarowinkler".equals(distanceVal)) {
        return new JaroWinklerDistance();
    } else if ("ngram".equals(distanceVal)) {
        return new NGramDistance();
    } else {/*  ww  w.  j  a va2s  . co  m*/
        throw new IllegalArgumentException("Illegal distance option " + distanceVal);
    }
}

---

From source file:org.elasticsearch.search.suggest.SuggestUtils.java

License:Apache License

public static StringDistance resolveDistance(String distanceVal) {
    if ("internal".equals(distanceVal)) {
        return DirectSpellChecker.INTERNAL_LEVENSHTEIN;
    } else if ("damerau_levenshtein".equals(distanceVal) || "damerauLevenshtein".equals(distanceVal)) {
        return new LuceneLevenshteinDistance();
    } else if ("levenstein".equals(distanceVal)) {
        return new LevensteinDistance();
        //TODO Jaro and Winkler are 2 people - so apply same naming logic as damerau_levenshtein  
    } else if ("jarowinkler".equals(distanceVal)) {
        return new JaroWinklerDistance();
    } else if ("ngram".equals(distanceVal)) {
        return new NGramDistance();
    } else {/*from  w  w  w.j av  a  2s. c  o  m*/
        throw new ElasticsearchIllegalArgumentException("Illegal distance option " + distanceVal);
    }
}

---

From source file:uk.ac.open.kmi.squire.ontologymatching.JaroWinklerSimilarity.java

@Override
public float computeMatchingScore(String s1, String s2) {
    JaroWinklerDistance d = new JaroWinklerDistance();
    return d.getDistance(s1, s2);
}

---