Examples of UniGram

Examples of seekfeel.miners.features.UniGram

21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
    public ArrayList<Feature> extractFeatures(CorpusHolder corpus) {

        ArrayList unigrams = new ArrayList<Feature>();
        ArrayList<DataUnit> positiveExamples = corpus.getPositiveExamples();
        ArrayList<DataUnit> negativeExamples = corpus.getNegativeExamples();
        UniGram tempUniGram;

        ArrayList<String> words;
        for (DataUnit posReview : positiveExamples) {
            words = TextSplitter.splitTextToWords(posReview.getDataBody());
            for (String word : words) {
                if (word.length() == 0) {
                    continue;
                }
                tempUniGram = new UniGram();
                tempUniGram.setUniGram(word);
                if (unigrams.contains(tempUniGram)) {
                    ((UniGram) unigrams.get(unigrams.indexOf(tempUniGram))).inCrementPositive();
                } else {
                    tempUniGram.setPositiveCount(1);
                    unigrams.add(tempUniGram);
                }
            }

        }
        for (DataUnit negReview : negativeExamples) {
            words = TextSplitter.splitTextToWords(negReview.getDataBody());
            for (String word : words) {
                if (word.length() == 0) {
                    continue;
                }
                tempUniGram = new UniGram();
                tempUniGram.setUniGram(word);
                if (unigrams.contains(tempUniGram)) {
                    ((UniGram) unigrams.get(unigrams.indexOf(tempUniGram))).inCrementNegative();
                } else {
                    tempUniGram.setNegativeCount(1);
                    unigrams.add(tempUniGram);
                }
            }

        }
View Full Code Here

Examples of seekfeel.miners.features.UniGram

64
65
66
67
68
69
70
71
72
73
74
75
76
77
    }

    @Override
    public ArrayList<Feature> filter(int countThreshold, ArrayList<Feature> allFeatures) {
        ArrayList<Feature> filteredFeatures = new ArrayList<Feature>();
        UniGram tempFeat;
        for (Feature feat : allFeatures) {
            tempFeat = (UniGram) feat;
            if ((tempFeat.getPositiveCount() + tempFeat.getNegativeCount()) >= countThreshold) {
                filteredFeatures.add(tempFeat);
            }
        }
        return filteredFeatures;
    }
View Full Code Here

Examples of seekfeel.miners.features.UniGram

22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
public class DeltaTFIDFComputerUni extends FeatureValueComputer {

    @Override
    public LinkedHashMap<Integer, Double> computeFeatures(DataUnit example, ArrayList<Feature> features, CorpusHolder corpus) {
        LinkedHashMap<Integer, Double> featuresValues = new LinkedHashMap<Integer, Double>();
        UniGram tempUnigram;
        // The Equation used in calculating feature value :
        // V = Count * log ( NumPositiveExamples * Count of term in Negative corpus / NumNegativeExamples * Count of term in positive corpus)
        double featureValue;
        int countInDoc = 0;
        double countInPos;
        double countInNeg;
        double posToNegRatio = (double) corpus.getNumPositives() / (double) corpus.getNumNegatives();
        HashMap<String, Integer> wordsWithCount = TextSplitter.getWordsCount(example.getDataBody());

        /*while (wordsIterator.hasNext()) {
            tempEnt = wordsIterator.next();
            featIndex = uniGrams.indexOf(tempEnt.getKey());
            if (featIndex >= 0) {
                tempUnigram = (UniGram) features.get(featIndex);
                countInDoc = tempEnt.getValue();
                countInPos = tempUnigram.getPositiveCount() + 0.0001;
                countInNeg = tempUnigram.getNegativeCount() + 0.0001;
                featureValue = countInDoc * (Math.log((posToNegRatio * (countInNeg / countInPos))));
                featuresValues.put(featIndex, featureValue);
            }
        }*/
        for (int i = 0; i < features.size(); i++) {
            tempUnigram = (UniGram) features.get(i);
            if (wordsWithCount.containsKey(((UniGram)features.get(i)).getUniGram())) {
                countInDoc = wordsWithCount.get(((UniGram)features.get(i)).getUniGram());
                countInPos = tempUnigram.getPositiveCount() + 0.0001;
                countInNeg = tempUnigram.getNegativeCount() + 0.0001;
                featureValue = countInDoc * (Math.log((posToNegRatio * (countInNeg / countInPos))));
                featuresValues.put(i, 1.0);
            }
        }
        return featuresValues;
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.