/**
* This function generaes smoothed unigramprobabilities for given data
*
* @param f
* @return
* @throws IOException
*/
public List<Gram> generateSmoothedProbabilities(File f) throws IOException {
Tokenizer tokenizer = new Tokenizer();
// Frequency map - this map contains frequency as the key and frequency
// of frequency as the value.
SmoothingInputBean bean = new SmoothingInputBean();
bean.setTokens(tokenizer.getTokens(f));
bean.setWordFrequencyMap(tokenizer.extractCorpusFrequency(bean.getTokens()));
bean.setDenominiator(bean.getTokens().getTokens().size());
List<Gram> grams = generateSmoothedProbabilities(bean);
return grams;
}
/*
* (non-Javadoc)
*
* @see hw1.service.NGramGenerator#generateProbabilities(java.io.File)
*/
public List<Gram> generateProbabilities(File f) throws IOException {
List<Gram> grams = new LinkedList<Gram>();
Tokenizer tokenizer = new Tokenizer();
TokenList tokens = tokenizer.getTokens(f);
Map<String, Integer> map = tokenizer.extractCorpusFrequency(tokens);
int count = 0;
for (Map.Entry<String, Integer> entry : map.entrySet()) {
count += entry.getValue();
}
for (Map.Entry<String, Integer> entry : map.entrySet()) {
grams.add(new Gram(entry.getKey(), (Utils.round(((double) entry.getValue() / count), 5))));
}
Collections.sort(grams);
return grams;
}
