public static void TestNames()
{
//def gender_features(word):
// return { 'last_letter': word[-1]}
dynamic GenderFeature(string word)
{
var dict = BuiltIns.Dict();
dict["last_letter"] = string.IsNullOrEmpty(word) ? "" : word.Last().ToString();
return(dict);
}
//from nltk.corpus import names
//>>> labeled_names = ([(name, 'male') for name in names.words('male.txt')] +
//... [(name, 'female') for name in names.words('female.txt')])
var namesCorpus = new NltkNet.Nltk.Corpus.Names();
var maleNames = namesCorpus.Words("male.txt").AsNet.Select(name => (name, gender: "male"));
var femaleNames = namesCorpus.Words("female.txt").AsNet.Select(name => (name, gender: "female"));
var labeledNames = maleNames.ToList();
labeledNames.AddRange(femaleNames);
//>>> import random
//>>> random.shuffle(labeled_names)
labeledNames.Shuffle();
//featuresets = [(gender_features(n), gender) for (n, gender) in labeled_names]
//train_set, test_set = featuresets[500:], featuresets[:500]
var featuresets = labeledNames.Select(nameAndGender => (GenderFeature(nameAndGender.name), nameAndGender.gender)).ToList();
var train_set = featuresets.SkipLast(500).ToList();
var test_set = featuresets.TakeLast(500).ToList();
// nltk.NaiveBayesClassifier.train(train_set)
dynamic classifier = Nltk.NaiveBayesClassifier.Train(train_set);
// classifier.classify(gender_features('Neo'))
Print("Neo=" + classifier.classify(GenderFeature("Neo")));
Print("Trinity=" + classifier.classify(GenderFeature("Trinity")));
Print(Nltk.Classify.Accuracy(classifier, test_set));
classifier.show_most_informative_features(5);
}
