lucenenet-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From geobmx540 <...@git.apache.org>
Subject [GitHub] lucenenet pull request: port of lucene-solr/lucene/classification ...
Date Sun, 07 Dec 2014 05:22:08 GMT
Github user geobmx540 commented on a diff in the pull request:

    https://github.com/apache/lucenenet/pull/26#discussion_r21422060
  
    --- Diff: src/Lucene.Net.Classification/KNearesteighborClassifier.cs ---
    @@ -0,0 +1,150 @@
    +/*
    + * Licensed to the Apache Software Foundation (ASF) under one or more
    + * contributor license agreements.  See the NOTICE file distributed with
    + * this work for additional information regarding copyright ownership.
    + * The ASF licenses this file to You under the Apache License, Version 2.0
    + * (the "License"); you may not use this file except in compliance with
    + * the License.  You may obtain a copy of the License at
    + *
    + *     http://www.apache.org/licenses/LICENSE-2.0
    + *
    + * Unless required by applicable law or agreed to in writing, software
    + * distributed under the License is distributed on an "AS IS" BASIS,
    + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    + * See the License for the specific language governing permissions and
    + * limitations under the License.
    + */
    +
    +namespace Lucene.Net.Classification
    +{
    +    using Lucene.Net.Analysis;
    +    using Lucene.Net.Index;
    +    using Lucene.Net.Queries.Mlt;
    +    using Lucene.Net.Search;
    +    using Lucene.Net.Util;
    +    using System;
    +    using System.Collections.Generic;
    +    using System.IO;
    +
    +    /// <summary>
    +    /// A k-Nearest Neighbor classifier (see <code>http://en.wikipedia.org/wiki/K-nearest_neighbors</code>)
based
    +    /// on {@link MoreLikeThis}
    +    /// 
    +    /// @lucene.experimental
    +    /// </summary>
    +    public class KNearestNeighborClassifier : Classifier<BytesRef> 
    +    {
    +
    +        private MoreLikeThis _mlt;
    +        private String[] _textFieldNames;
    +        private String _classFieldName;
    +        private IndexSearcher _indexSearcher;
    +        private readonly int _k;
    +        private Query _query;
    +
    +        private int _minDocsFreq;
    +        private int _minTermFreq;
    +
    +        /// <summary>Create a {@link Classifier} using kNN algorithm</summary>
    +        /// <param name="k">the number of neighbors to analyze as an <code>int</code></param>
    +        public KNearestNeighborClassifier(int k) 
    +        {
    +        this._k = k;
    +        }
    +
    +        /// <summary>Create a {@link Classifier} using kNN algorithm</summary>
    +        /// <param name="k">the number of neighbors to analyze as an <code>int</code></param>
    +        /// <param name="minDocsFreq">the minimum number of docs frequency for
MLT to be set with {@link MoreLikeThis#setMinDocFreq(int)}</param>
    +        /// <param name="minTermFreq">the minimum number of term frequency for
MLT to be set with {@link MoreLikeThis#setMinTermFreq(int)}</param>
    +        public KNearestNeighborClassifier(int k, int minDocsFreq, int minTermFreq) 
    +        {
    +        this._k = k;
    +        this._minDocsFreq = minDocsFreq;
    +        this._minTermFreq = minTermFreq;
    +        }
    +
    +        public ClassificationResult<BytesRef> AssignClass(String text)
    +        {
    +            if (_mlt == null) 
    +            {
    +                throw new IOException("You must first call Classifier#train");
    +            }
    +
    +            BooleanQuery mltQuery = new BooleanQuery();
    +            foreach (String textFieldName in _textFieldNames) 
    +            {
    +                mltQuery.Add(new BooleanClause(_mlt.Like(new StringReader(text), textFieldName),
BooleanClause.Occur.SHOULD));
    +            }
    +            Query classFieldQuery = new WildcardQuery(new Term(_classFieldName, "*"));
    +            mltQuery.Add(new BooleanClause(classFieldQuery, BooleanClause.Occur.MUST));
    +            if (_query != null) {
    +                mltQuery.Add(_query, BooleanClause.Occur.MUST);
    +            }
    +            TopDocs topDocs = _indexSearcher.Search(mltQuery, _k);
    +            return SelectClassFromNeighbors(topDocs);
    +        }
    +
    +        private ClassificationResult<BytesRef> SelectClassFromNeighbors(TopDocs
topDocs) 
    +        {
    +            // TODO : improve the nearest neighbor selection
    +            Dictionary<BytesRef, int> classCounts = new Dictionary<BytesRef,
int>();
    +
    +            foreach (ScoreDoc scoreDoc in topDocs.ScoreDocs) 
    +            {
    +                BytesRef cl = new BytesRef(_indexSearcher.Doc(scoreDoc.Doc).GetField(_classFieldName).StringValue);
    +                int count = classCounts[cl];
    +                if (classCounts.ContainsKey(cl))
    +                {
    +                    classCounts[cl] = count + 1;
    +                } 
    +                else 
    +                {
    +                    classCounts.Add(cl, 1);
    +                }
    +            }
    +            double max = 0;
    +            BytesRef assignedClass = new BytesRef();
    +            foreach (KeyValuePair<BytesRef, int> entry in classCounts) 
    +            {
    +                int count = entry.Value;
    +                if (count > max) 
    +                {
    +                    max = count;
    +                    assignedClass = (BytesRef)entry.Key.Clone();
    +                }
    +            }
    +            double score = max / (double) _k;
    +            return new ClassificationResult<BytesRef>(assignedClass, score);
    +        }
    +
    +        public void Train(AtomicReader atomicReader, String textFieldName, String classFieldName,
Analyzer analyzer) 
    +        {
    +            Train(atomicReader, textFieldName, classFieldName, analyzer, null);
    +        }
    +
    +
    +        public void Train(AtomicReader atomicReader, String textFieldName, String classFieldName,
Analyzer analyzer, Query query) 
    +        {
    +            Train(atomicReader, new String[]{textFieldName}, classFieldName, analyzer,
query);
    +        }
    +
    +        public void Train(AtomicReader atomicReader, String[] textFieldNames, String
classFieldName, Analyzer analyzer, Query query)
    +        {
    +            this._textFieldNames = textFieldNames;
    --- End diff --
    
    could probably get rid of the references to `this.` since the private members are prefixed
with an underscore


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastructure@apache.org or file a JIRA ticket
with INFRA.
---

Mime
View raw message