public int Classify(string review)
{
var tokenizer = new HappyFunTokenizer(_preserveCase);
var tokens = this.AddNegationAugments(tokenizer.Tokenize(review));
var scores = new Dictionary <int, double>();
for (var i = 1; i < 6; i++)
{
scores[i] = this.GetEmptyScoreForSentiment(i);
}
foreach (var token in tokens)
{
if (!Utilities.CheckUtf8(token))
{
continue;
}
if (!this._wordList.ContainsKey(token))
{
continue;
}
var index = this._wordList[token];
for (var i = 1; i < 6; i++)
{
var score = this.GetProbabilityOfWordGivenSentimentFast(index, i);
scores[i] *= score / (1.0 - score);
}
}
return(scores.Aggregate((l, r) => l.Value > r.Value ? l : r).Key);
}
protected int AddTrainingEntry(HappyFunTokenizer tokenizer, int score, string review)
{
var addedEntries = new HashSet <int>();
var tokens = this.AddNegationAugments(tokenizer.Tokenize(review));
var count = 0;
foreach (var token in tokens)
{
if (!Utilities.CheckUtf8(token))
{
continue;
}
if (this._wordList.ContainsKey(token))
{
var index = this._wordList[token];
if (addedEntries.Contains(index))
{
continue;
}
count++;
this._sentimentWordCounts[score][index]++;
addedEntries.Add(index);
}
else
{
var index = this._wordList.Count;
this._wordList[token] = index;
count++;
for (var i = 0; i < 6; i++)
{
this._sentimentWordCounts[i][index] = 0;
}
this._sentimentWordCounts[score][index]++;
addedEntries.Add(index);
}
}
return(count);
}