before presentation

classifier.py

@@ -0,0 +1,172 @@
+#! /usr/bin/env python
+# -*- coding: UTF-8 -*-
+
+"""
+Classifier
+==========
+After Feature Extraction, that returns a data of the format
+[(filename, linenum, vote, sentence, feat1, feat2, ...)]
+"""
+
+from __future__ import division
+from optparse import OptionParser
+import json
+from pprint import pprint
+import random
+import math
+import nltk
+
+def getUserInput():
+    optionparser = OptionParser()
+
+    optionparser.add_option('-i', '--input', dest='inputfile')
+
+
+    (option, args) = optionparser.parse_args()
+
+    if not option.inputfile:
+        return optionparser.error('html file input not provided.\n Usage: --url="path.to.appurl"')
+
+    return { 'file' : option.inputfile }
+
+
+
+def fileExtractor(f):
+    fObj = open(f)
+    data = json.load(fObj)
+
+    fObj.close()
+
+    return data
+
+
+def featureAggregator(extract):
+    outputdata = []
+
+    for app in extract:
+        outputdata.append(featureExtractor(app))
+
+    return outputdata
+
+
+def featureExtractor(app):
+    featDict = {}
+
+    # featList['price'] = getAppPrice(app)
+    # featList['numrev'] = getNumReviews(app)
+    featDict['1starrating'] = getOneStarRating(app)
+    featDict['2starrating'] = getTwoStarRating(app)
+    featDict['3starRating'] = getThreeStarRating(app)
+    featDict['4starRating'] = getFourStarRating(app)
+    featDict['5starRating'] = getFiveStarRating(app)
+    featDict['hasPrivacy'] = getPrivacyState(app)
+
+    return featDict
+
+
+
+
+def getAppPrice(app):
+    return app['price']
+
+def getNumReviews(app):
+    return len(app['reviews'])
+
+
+def getOneStarRating(app):
+    for appRatingCount in app['rating']:
+        if appRatingCount[0] == ' 1 ':
+            return appRatingCount[1]
+
+
+def getTwoStarRating(app):
+    for appRatingCount in app['rating']:
+        if appRatingCount[0] == ' 2 ':
+            return appRatingCount[1]
+
+
+def getThreeStarRating(app):
+    for appRatingCount in app['rating']:
+        if appRatingCount[0] == ' 3 ':
+            return appRatingCount[1]
+
+def getFourStarRating(app):
+    for appRatingCount in app['rating']:
+        if appRatingCount[0] == ' 4 ':
+            return appRatingCount[1]
+
+
+def getFiveStarRating(app):
+    for appRatingCount in app['rating']:
+        if appRatingCount[0] == ' 5 ':
+            return appRatingCount[1]
+
+
+
+def getPrivacyState(app):
+    if app['devprivacyurl'] == 'N.A':
+        return False
+    else:
+        return True
+
+
+def classifier(extract, fold=10):
+
+    labeldata = 'fair'
+
+    data = []
+    for app in extract:
+        for rev in app['reviews']:
+            revlower = rev[1].lower()
+
+            # print "reviews" , revlower
+            if revlower.find('fake') != -1:
+
+                labeldata = 'unfair'
+
+        features = featureExtractor(app)
+
+        data.append([labeldata, list(features.values())])
+
+
+    # pprint(data)
+
+    random.shuffle(data)
+
+    claccuracy = []
+    size = int(math.floor(len(data) / 10.0))
+
+    for i in range(fold):
+        test_this_round = data[i*size:][:size]
+        train_this_round = data[:i*size] + data[(i+1)*size:]
+
+        acc = myclassifier(train_this_round, test_this_round)
+
+        claccuracy.append(acc)
+
+
+
+def myclassifier(train_data, test_data):
+    classifier = nltk.NaiveBayesClassifier.train(train_data)
+
+
+    # print classifier.show_most_informative_features()
+    return nltk.classify.accuracy(classifier, test_data)
+
+
+
+
+
+
+
+def main():
+    userinput = getUserInput()
+    print userinput['file']
+
+    extract = fileExtractor(userinput['file'])
+    # features = featureAggregator(extract)
+    classifier(extract)
+
+
+if __name__ == "__main__":
+    main()

exports/game_brain_all.json

@@ -5617,4 +5617,5 @@
         "totalReviewers": 10576,
         "version": "1.2"
     }
+
 ]