public training(int lang_id, int category_id, string langauge, string stemming, bool addtur, int n, bool punctuation, int nov) { _cr = new classifier(langauge, stemming, addtur, n, punctuation, nov);//classifier hazır _category_sentence_count = new Dictionary <int, int>(); all_sentence_count = 0; word_dic _wd = new word_dic(); comment _c = new comment(); DataTable _dt = _c.select_comment(lang_id.ToString(), category_id.ToString()); int id = 1; int yuzde50 = _dt.Rows.Count / 2; foreach (DataRow row in _dt.Rows) // Loop over the rows. { var type = row[0]; string _sentence = row[1].ToString().ToLower(); int _type = Convert.ToInt32(type); _cr.AddSentenceforTraining(_sentence, _type); AddASentence(_type);//kategorideki cümle sayısı all_sentence_count++; id++; if (id == yuzde50) { break; } } }
public arff_converter2(int lang_id, int category_id, string langauge, string stemming, bool addtur, int n, bool punctuation, int nov) { word_dic _wd = new word_dic(); comment _c = new comment(); DataTable _dt = _c.select_comment(lang_id.ToString(), category_id.ToString()); int olumsuz = 0; int olumlu = 0; int notr = 0; int soru = 0; int id = 1; foreach (DataRow row in _dt.Rows) // Loop over the rows. { var type = row[0]; string _sentence = row[1].ToString().ToLower(); int _type = Convert.ToInt32(type); if (_type == -1) { olumsuz++; if (olumsuz < 3300000) { _wd.update_hashtables(_sentence, _type, langauge, stemming, addtur, n, punctuation, nov); } } if (_type == 0) { notr++; if (notr < 1700000) { _wd.update_hashtables(_sentence, _type, langauge, stemming, addtur, n, punctuation, nov); } } if (_type == 1) { olumlu++; if (olumlu < 550000000) { _wd.update_hashtables(_sentence, _type, langauge, stemming, addtur, n, punctuation, nov); } } if (_type == 2) { soru++; if (soru < 100000) { _wd.update_hashtables(_sentence, _type, langauge, stemming, addtur, n, punctuation, nov); } } id++; } _sentence_data = _wd._sentence_data; }
public int[,] ConfusionMatrix(int lang_id, int category_id) { int[,] _cm = new int[_cat_dic.Count, _cat_dic.Count]; for (int x = 0; x < _cat_dic.Count; x++) { for (int y = 0; y < _cat_dic.Count; y++) { _cm[x, y] = 0; } } Hashtable _categoriler = new Hashtable(); int i = 0; foreach (var _cat in _cat_dic) { _categoriler.Add((int)_cat.Key, i); i++; } word_dic _wd = new word_dic(); comment _c = new comment(); DataTable _dt = _c.select_comment(lang_id.ToString(), category_id.ToString()); int id = 1; foreach (DataRow row in _dt.Rows) // Loop over the rows. { var type = row[0]; string _sentence = row[1].ToString().ToLower(); int _type = Convert.ToInt32(type); int x = (int)_categoriler[_type]; int y = (int)_categoriler[Estimate_Category(_sentence)]; _cm[x, y]++; id++; } return(_cm); }
public Hashtable _words_in_sentence;//bir cümle içindeki kelimeler için public arff_converter(int lang_id, int category_id, string langauge, string stemming, bool addtur, int n, bool punctuation, int nov) { word_dic _wd = new word_dic(); comment _c = new comment(); DataTable _dt = _c.select_comment(lang_id.ToString(), category_id.ToString()); int id = 1; foreach (DataRow row in _dt.Rows) // Loop over the rows. { var type = row[0]; string _sentence = row[1].ToString().ToLower(); int _type = Convert.ToInt32(type); _wd.update_hashtables(_sentence, _type, langauge, stemming, addtur, n, punctuation, nov); id++; } _word_dic = _wd._word_dic; _sentence_word = _wd._sentence_word; _sentence_type = _wd._sentence_type; }