/// <summary>
/// 训练现有样本,返回d和total
/// </summary>
/// <param name="negFilePath">负面词训练集路径</param>
/// <param name="negWords">负面词库文本</param>
/// <param name="posFilePath">正面词训练集路径</param>
/// <param name="posWords">正面词库文本</param>
/// <param name="d">存储正面和负面词集的字典</param>
/// <param name="total">总数</param>
/// <param name="stopwordFilepath">排除词路径</param>
public static void Train_data(string negFilePath, string negWords, string posFilePath, string posWords,
ref Dictionary <string, AddOneProb> d, ref double total, string stopwordFilepath)
{
//d = new Dictionary<string, AddOneProb>() { { "pos", new AddOneProb() }, { "neg", new AddOneProb() } };
string negfile = "", posfile = "";
using (var sr1 = new StreamReader(negFilePath, Encoding.Default))
negfile = sr1.ReadToEnd();
using (var sr2 = new StreamReader(posFilePath, Encoding.Default))
posfile = sr2.ReadToEnd();
string stopwords = ReadTxtToEnd(stopwordFilepath);
List <Tuple <List <string>, string> > data = new List <Tuple <List <string>, string> >();
var sent_cut = new Jieba();
sent_cut.NegWords = negWords;
sent_cut.PosWords = posWords;
foreach (var sent in posfile.Replace("\r", "").Split('\n'))
{
sent_cut.doc = sent;
sent_cut.stopwords = stopwords;
//<Question>why not work
//var data_pos = new Tuple<List<string>,string>();
//</Question>
data.Add(new Tuple <List <string>, string>(sent_cut.handle_sentiment(), "pos"));
}
Console.WriteLine("正面词库导入完毕");
foreach (var sent in negfile.Replace("\r", "").Split('\n'))
{
sent_cut.doc = sent;
sent_cut.stopwords = stopwords;
data.Add(new Tuple <List <string>, string>(sent_cut.handle_sentiment(false), "neg"));
}
Console.WriteLine("负面词库导入完毕");
foreach (var d_ in data)
{
var c = d_.Item2.ToString();
if (d_.Item1 == null)
{
continue;
}
else
{
foreach (var word in d_.Item1)
{
d[c].add(word, 1);
}
}
}
///<question>字典所有值求和
//d.Sum(x=>d[x])
///</question>
total = 0;
foreach (var value in d.Values)
{
total += value.total;
}
}
public static double classify_(string sent, Dictionary <string, AddOneProb> d,
double total, string stopwordFilepath)
{
Jieba jiebaword = new Jieba();
jiebaword.doc = sent;
jiebaword.stopwords = ReadTxtToEnd(stopwordFilepath);
var retprob = Classify(jiebaword.JiebaCut(), d, total);
if (retprob.Item1 == "pos")
{
return(retprob.Item2);
}
else
{
return(1 - retprob.Item2);
}
}
