static List <string> slipt(string n)
{
NLP nlp = new NLP();
char[] delimate = new char[] { '\'', ':', ',', '.', '(', ')', '/', '\"', '!', '*', ';', '[', ']', '{', '}' };
List <string> temp = new List <string>();
/*sentence detect*/
string[] sentences = nlp.SentDetect(n);
foreach (string s in sentences)
{
/*tokenization*/
string[] tokens2 = nlp.Tokenize(s);
/*Stemming, Lemmatization*/
for (int i = 0; i < tokens2.Length; i++)
{
tokens2[i] = nlp.Lemmatization(tokens2[i]);
}
/*Filter out stopwords*/
string[] result2 = nlp.FilterOutStopWords(tokens2);
foreach (string sf in result2)
{
string[] te = sf.Split(delimate);
foreach (string t in te)
{
if (t != "")
{
temp.Add(t);
}
}
}
}
return(temp);
/*Paser
* Parse p = nlp.Parser(sent);
* p.show();*/
//Console.ReadKey();
}
static void Main(string[] args)
{
string r_path = "./input_hotel_review.txt";
string w_path = "./output.txt";
StreamWriter sw = new StreamWriter(w_path);
StreamReader sr = new StreamReader(r_path);
NLP nlp = new NLP();
while (!sr.EndOfStream)
{
int index = 0, count = 0;
List <indexcoll> option_index = new List <indexcoll>();
string str = sr.ReadLine();
do
{
index = str.IndexOf("<opinion>", index);
if (index != -1)
{
int end_index = str.IndexOf("</opinion>", index);
indexcoll temp = new indexcoll();
temp.start = index - (19 * count);
index += 9;
temp.end = end_index - index;
option_index.Add(temp);
count++;
}
} while (index != -1);
Console.WriteLine(str);
index = count = 0;
//chinese word segmentation
string[] result = nlp.CWS(str);
foreach (string s in result)
{
string[] tokens = s.Split(' ');
foreach (string t in tokens)
{
char[] separators = { '(', ')' };
string[] temp = t.Split(separators);
for (int i = 0, flag = 0; i < temp[0].Length; i++, count++)
{
//[char][\t][POS][\t][斷詞後的長度][\t][斷詞的B,I][\t][Ans]
sw.Write(temp[0][i] + "\t" + temp[1] + "\t" + temp[0].Length + "\t");
if (i == 0)
{
sw.Write("B\t");
if (index < option_index.Count() && option_index[index].start.Equals(count))
{
sw.Write("B-OPINION\t");
flag = 1;
}
else
{
sw.Write("0\t");
if (flag == 1)
{
flag = 0;
index++;
}
}
}
else
{
sw.Write("I\t");
if (flag == 1 && (option_index[index].start + option_index[index].end) > count)
{
sw.Write("I-OPINION\t");
}
else
{
sw.Write("0\t");
}
}
sw.WriteLine();
}
}
}
sw.WriteLine();
}
}
public void Nlp(string folderName)
{
NLP nlp = new NLP();
StreamReader srP = new StreamReader(@"C:\Users\Ian Hsieh\Downloads\Project Testsite\SenticDic\positive-words.txt");
StreamReader srN = new StreamReader(@"C:\Users\Ian Hsieh\Downloads\Project Testsite\SenticDic\negative-words.txt");
StreamReader srGA = new StreamReader(@"C:\Users\Ian Hsieh\Downloads\Project Testsite\Gold_answer_test.txt");
StreamReader srSN = new StreamReader(@"C:\Users\Ian Hsieh\Downloads\Project Testsite\Songnames_test.txt");
string[] pw = new string[2048];
string[] nw = new string[5000];
Dictionary <string, string> answer = new Dictionary <string, string>();
int i = 0;
string line, line1;
while ((line = srP.ReadLine()) != null)
{
pw[i++] = nlp.Lemmatization(nlp.Stem(line));
}
i = 0;
while ((line = srN.ReadLine()) != null)
{
nw[i++] = nlp.Lemmatization(nlp.Stem(line));
}
i = 0;
while ((line = srGA.ReadLine()) != null && (line1 = srSN.ReadLine()) != null)
{
//if (line == "1" || line == "4")
//answer.Add(line1.Replace('?', ' ').Replace(':', ' '), "1");
//else if (line == "3" || line == "2")
answer.Add(line1.Replace('?', ' ').Replace(':', ' '), line);
}
srP.Close();
srN.Close();
srGA.Close();
srSN.Close();
StreamWriter sw = new StreamWriter(@"C:\Users\Ian Hsieh\Downloads\Project Testsite\MIR_TF_test.txt");
int number = 1, tv, no;
Dictionary <string, int> dict = new Dictionary <string, int>();
Dictionary <string, int> dic = new Dictionary <string, int>();
List <string> avg = new List <string>();
foreach (string fileName in System.IO.Directory.GetFiles(folderName))
{
System.IO.StreamReader file = new System.IO.StreamReader(fileName);
while ((line = file.ReadLine()) != null)
{
string[] sents = nlp.SentDetect(line.Trim());
string[] tokens, tokens_reviews;
foreach (string sent in sents)
{
if (sent.Contains('('))
{
sent.Replace('(', ' ');
}
if (sent.Contains(')'))
{
sent.Replace(')', ' ');
}
if (sent.Contains('!'))
{
sent.Replace('!', ' ');
}
if (sent.Contains('#'))
{
sent.Replace('#', ' ');
}
if (sent.Contains('&'))
{
sent.Replace('&', ' ');
}
if (sent.Contains('*'))
{
sent.Replace('*', ' ');
}
if (sent.Contains(','))
{
sent.Replace(',', ' ');
}
if (sent.Contains('.'))
{
sent.Replace('.', ' ');
}
if (sent.Contains(':'))
{
sent.Replace(':', ' ');
}
if (sent.Contains(';'))
{
sent.Replace(';', ' ');
}
if (sent.Contains('?'))
{
sent.Replace('?', ' ');
}
if (sent.Contains('"'))
{
sent.Replace('"', ' ');
}
if (sent.Contains('\\'))
{
sent.Replace('\\', ' ');
}
if (sent.Contains('/'))
{
sent.Replace('/', ' ');
}
if (sent.Contains('-'))
{
sent.Replace('-', ' ');
}
if (sent == " ")
{
continue;
}
tokens = nlp.Tokenize(sent);//tokenize sentences
tokens_reviews = nlp.Lemmatization(nlp.Stem(nlp.FilterOutStopWords(tokens)));
foreach (string token in tokens_reviews)
{
if (dict.ContainsKey(token))
{
dict[token]++;
}
else
{
dict.Add(token, 1);
}
if (dic.ContainsKey(token))
{
continue;
}
else
{
dic.Add(token, number++);
}
}
}
}
/*if(System.IO.Directory.GetFiles(folderName).Last() == fname)
* file.Close();*/
sw.Write("+" + answer[fileName.Substring(fileName.IndexOf('_') + 1, (fileName.IndexOf(".txt") - fileName.IndexOf('_') - 1))] + " ");
tv = 0;
no = 0;
foreach (KeyValuePair <string, int> item in dict)
{
int value = 1;//0;
/*foreach (string word in pw)
* {
* if (word == item.Key)
* {
* value = 1;
* break;
* }
* }
* foreach(string word in nw)
* {
* if (word == item.Key)
* {
* value = -1;
* break;
* }
* }
* if (value == 0 && !avg.Contains(item.Key))
* {
* avg.Add(item.Key);
* continue;
* }*/
Console.WriteLine(dic[item.Key] + ":" + (item.Value * value) + "\n");
sw.Write(dic[item.Key] + ":" + (item.Value * value) + " ");
tv += item.Value * value;
no += item.Value;
}
/*foreach (string str in avg)
* {
* Console.WriteLine(dic[str] + ":" + ((double)dict[str] * (double)tv / (double)no) + "\n");
* sw.Write(dic[str] + ":" + ((double)dict[str] * (double)tv / (double)no) + " ");
* }*/
sw.WriteLine();
dict.Clear();
avg.Clear();
}
sw.Close();
}
static void Main(string[] args)
{
NLP nlp = new NLP();
XmlDocument doc = new XmlDocument();
doc.Load("training.xml");
XmlNodeList nodes = doc.SelectNodes("RDF/Text");
Dictionary <string, int> dict = new Dictionary <string, int>();
char[] separator = { ' ', ',', '.', '?', '!' };
int tokno = 1;
StreamWriter sw = new StreamWriter("./Task1op.txt");
foreach (XmlNode n in nodes)
{
string text = n.InnerText;
string[] sents = nlp.SentDetect(text);
string p = n.Attributes["category"].Value;
int polar = 0;
if (p == "book")
{
polar = 1;
}
else if (p == "dvd")
{
polar = 2;
}
else if (p == "health")
{
polar = 3;
}
else if (p == "music")
{
polar = 4;
}
else if (p == "toys_games")
{
polar = 5;
}
sw.Write("+" + polar.ToString() + " ");
sw.Flush();
foreach (string sent in sents)
{
string s = sent.ToLower();
string[] token = nlp.Tokenize(s);
foreach (string t in token)
{
string lemma = nlp.Lemmatization(t);
if (nlp.IsStopWord(lemma))
{
continue;
}
else
{
if (!dict.ContainsKey(lemma))
{
dict.Add(lemma, tokno);
tokno++;
}
sw.Write(dict[lemma] + ":1 ");
sw.Flush();
}
}
}
sw.WriteLine("");
sw.Flush();
}
Console.WriteLine("Finished.");
Console.ReadKey();
}