public WordProbability GetWordProbability(string category, string word, int? catId = null)
{
try
{
WordProbability wp = null;
long matchingCount = 0;
long nonMatchingCount = 0;
WordSource source = null;
if (catId != null)
{
source = this.Data.WordSources.All(new string[] { "Category" }).FirstOrDefault(w => w.Word == word && w.CategoryId == catId.Value);
}
else
{
source = this.Data.WordSources.All(new string[] { "Category" }).FirstOrDefault(w => w.Word == word && w.Category.Name == category);
}
if (source != null)
{
matchingCount = source.Matches;
nonMatchingCount = source.NonMatches;
}
wp = new WordProbability(word, matchingCount, nonMatchingCount);
return wp;
}
catch (Exception ex)
{
throw new WordsDataSourceException("Problem updating WordProbability.", ex);
}
}
public double CalculateOverallProbability(WordProbability[] wps)
{
if (wps == null || wps.Length == 0)
return IClassifierConstants.NEUTRAL_PROBABILITY;
// we need to calculate xy/(xy + z) where z = (1 - x)(1 - y)
// first calculate z and xy
double z = 0d;
double xy = 0d;
for (int i = 0; i < wps.Length; i++)
{
if (z == 0)
z = (1 - wps[i].Probability);
else
z = z * (1 - wps[i].Probability);
if (xy == 0)
xy = wps[i].Probability;
else
xy = xy * wps[i].Probability;
}
double numerator = xy;
double denominator = xy + z;
return numerator / denominator;
}
