public void TestCaseSensitive()
{
var classifier = new BayesianClassifier();
Assert.IsFalse(classifier.IsCaseSensitive);
classifier.IsCaseSensitive = true;
Assert.IsTrue(classifier.IsCaseSensitive);
}
private void CalculateProbability()
{
_probability = IClassifierConstants.NEUTRAL_PROBABILITY;
if (_matchingCount == 0)
{
_probability = _nonMatchingCount == 0 ? IClassifierConstants.NEUTRAL_PROBABILITY : IClassifierConstants.LOWER_BOUND;
}
else
{
_probability = BayesianClassifier.NormalizeSignificance(_matchingCount / (double)(_matchingCount + _nonMatchingCount));
}
}
public void TestClassify()
{
var wds = new SimpleWordsDataSource();
var classifier = new BayesianClassifier(wds);
var sentence = new[] { "This", "is", "a", "sentence", "about", "java" };
Assert.AreEqual(IClassifierConstants.NEUTRAL_PROBABILITY, classifier.Classify(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence), 0d);
wds.SetWordProbability(new WordProbability("This", .5d));
wds.SetWordProbability(new WordProbability("is", .5d));
wds.SetWordProbability(new WordProbability("a", .5d));
wds.SetWordProbability(new WordProbability("sentence", .2d));
wds.SetWordProbability(new WordProbability("about", .5d));
wds.SetWordProbability(new WordProbability("java", .99d));
Assert.AreEqual(.96d, classifier.Classify(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence), .009d);
}
public void TestCalculateOverallProbability()
{
var prob = 0.3d;
var wp1 = new WordProbability("myWord1", prob);
var wp2 = new WordProbability("myWord2", prob);
var wp3 = new WordProbability("myWord3", prob);
var wps = new[] { wp1, wp2, wp3 };
var errorMargin = 0.0001d;
var xy = (prob * prob * prob);
var z = (1-prob)*(1-prob)*(1-prob);
var result = xy/(xy + z);
var classifier = new BayesianClassifier();
Assert.AreEqual(result, classifier.CalculateOverallProbability(wps), errorMargin);
}
public void TestCalculateOverallProbability()
{
double prob = 0.3d;
WordProbability wp1 = new WordProbability("myWord1", prob);
WordProbability wp2 = new WordProbability("myWord2", prob);
WordProbability wp3 = new WordProbability("myWord3", prob);
WordProbability[] wps = new WordProbability[] { wp1, wp2, wp3 };
double errorMargin = 0.0001d;
double xy = (prob * prob * prob);
double z = (1-prob)*(1-prob)*(1-prob);
double result = xy/(xy + z);
BayesianClassifier classifier = new BayesianClassifier();
Assert.AreEqual(result, classifier.CalculateOverallProbability(wps), errorMargin);
}
private void CalculateProbability()
{
double result = IClassifierConstants.NEUTRAL_PROBABILITY;
if (_matchingCount == 0)
{
if (_nonMatchingCount == 0)
{
result = IClassifierConstants.NEUTRAL_PROBABILITY;
}
else
{
result = IClassifierConstants.LOWER_BOUND;
}
}
else
{
result = BayesianClassifier.NormalizeSignificance((double)_matchingCount / (double)(_matchingCount + _nonMatchingCount));
}
_probability = result;
}
public void TestTeaching()
{
BayesianClassifier classifier = new BayesianClassifier();
string[] sentence1 = new string[] {"The", "menu", "tag", "library", "manages", "the",
"complex", "process", "of", "creating", "menus", "in",
"JavaScript", "The", "menu", "tag", "itself", "is",
"an", "abstract", "class", "that", "extends", "the",
"TagSupport", "class", "and", "overrides", "the",
"doStartTag", "and", "doEndTag", "methods.", "The",
"getMenu", "method,", "which", "is", "a", "template",
"method", "and", "should", "be", "overridden", "in",
"the", "subclasses,", "provides", "JavaScript", "to",
"add", "menu", "items", "in", "the", "menu",
"structure", "created", "in", "the", "doStartTag",
"method", "Subclasses", "of", "the", "menu", "tag",
"override", "the", "getMenu", "method,", "which",
"uses", "menu", "builders", "to", "render", "menu",
"data", "from", "the", "data", "source"};
string[] sentence2 = new string[] {"I", "witness", "a", "more", "subtle",
"demonstration", "of", "real", "time", "physics",
"simulation", "at", "the", "tiny", "Palo", "Alto",
"office", "of", "Havok", "a", "competing", "physics",
"engine", "shop", "On", "the", "screen", "a",
"computer", "generated", "sailboat", "floats", "in",
"a", "stone", "lined", "pool", "of", "water", "The",
"company's", "genial", "Irish", "born", "cofounder",
"Hugh", "Reynolds", "shows", "me", "how", "to",
"push", "the", "boat", "with", "a", "mouse", "When",
"I", "nudge", "it", "air", "fills", "the", "sail",
"causing", "the", "ship", "to", "tilt", "leeward",
"Ripples", "in", "the", "water", "deflect", "off",
"the", "stones", "intersecting", "with", "one",
"another", "I", "urge", "the", "boat", "onward",
"and", "it", "glides", "effortlessly", "into", "the",
"wall", "Reynolds", "tosses", "in", "a", "handful",
"of", "virtual", "coins", "they", "spin", "through",
"the", "air,", "splash", "into", "the", "water,",
"and", "sink"};
string[] sentence3 = new string[] {"The", "New", "Input", "Output", "NIO", "libraries",
"introduced", "in", "Java", "2", "Platform",
"Standard", "Edition", "J2SE", "1.4", "address",
"this", "problem", "NIO", "uses", "a", "buffer",
"oriented", "model", "That", "is", "NIO", "deals",
"with", "data", "primarily", "in", "large", "blocks",
"This", "eliminates", "the", "overhead", "caused",
"by", "the", "stream", "model", "and", "even", "makes",
"use", "of", "OS", "level", "facilities", "where",
"possible", "to", "maximize", "throughput"};
string[] sentence4 = new string[] {"As", "governments", "scramble", "to", "contain",
"SARS", "the", "World", "Health", "Organisation",
"said", "it", "was", "extending", "the", "scope", "of",
"its", "April", "2", "travel", "alert", "to",
"include", "Beijing", "and", "the", "northern",
"Chinese", "province", "of", "Shanxi", "together",
"with", "Toronto", "the", "epicentre", "of", "the",
"SARS", "outbreak", "in", "Canada"};
string[] sentence5 = new string[] {"That", "was", "our", "worst", "problem", "I",
"tried", "to", "see", "it", "the", "XP", "way", "Well",
"what", "we", "can", "do", "is", "implement",
"something", "I", "can't", "give", "any", "guarantees",
"as", "to", "how", "much", "of", "it", "will", "be",
"implemented", "in", "a", "month", "I", "won't",
"even", "hazard", "a", "guess", "as", "to", "how",
"long", "it", "would", "take", "to", "implement", "as",
"a", "whole", "I", "can't", "draw", "UML", "diagrams",
"for", "it", "or", "write", "technical", "specs",
"that", "would", "take", "time", "from", "coding",
"it", "which", "we", "can't", "afford", "Oh", "and",
"I", "have", "two", "kids", "I", "can't", "do", "much",
"OverTime", "But", "I", "should", "be", "able", "to",
"do", "something", "simple", "that", "will", "have",
"very", "few", "bugs", "and", "show", "a", "working",
"program", "early", "and", "often"};
classifier.TeachMatch(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence1);
classifier.TeachNonMatch(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence2);
classifier.TeachMatch(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence3);
classifier.TeachNonMatch(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence4);
classifier.TeachMatch(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence5);
Assert.IsTrue(classifier.IsMatch(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence1));
Assert.IsTrue(!classifier.IsMatch(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence2));
Assert.IsTrue(classifier.IsMatch(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence3));
Assert.IsTrue(!classifier.IsMatch(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence4));
Assert.IsTrue(classifier.IsMatch(ICategorizedClassifierConstants.DEFAULT_CATEGORY, sentence5));
}
public void TestGetStopWordProvider()
{
var wds = new SimpleWordsDataSource();
ITokenizer tokenizer = new DefaultTokenizer(DefaultTokenizer.BREAK_ON_WORD_BREAKS);
IStopWordProvider stopWordProvider = new DefaultStopWordProvider();
var classifier = new BayesianClassifier(wds, tokenizer, stopWordProvider);
Assert.AreEqual(stopWordProvider, classifier.StopWordProvider);
}
public void TestTransformWord()
{
var classifier = new BayesianClassifier();
Assert.IsFalse(classifier.IsCaseSensitive);
string word = null;
try
{
classifier.TransformWord(word);
Assert.Fail("No exception thrown when null passed.");
}
catch {}
word = "myWord";
Assert.AreEqual(word.ToLower(), classifier.TransformWord(word));
classifier.IsCaseSensitive = true;
Assert.AreEqual(word, classifier.TransformWord(word));
}
public void TestGetWordsDataSource()
{
var wds = new SimpleWordsDataSource();
var classifier = new BayesianClassifier(wds);
Assert.AreEqual(wds, classifier.WordsDataSource);
}
public void TestGetTokenizer()
{
SimpleWordsDataSource wds = new SimpleWordsDataSource();
ITokenizer tokenizer = new DefaultTokenizer(DefaultTokenizer.BREAK_ON_WORD_BREAKS);
BayesianClassifier classifier = new BayesianClassifier(wds, tokenizer);
Assert.AreEqual(tokenizer, classifier.Tokenizer);
}
