private List <TaggedWord> JoinVerbParts(List <TaggedWord> sentence)
{
sentence.Reverse();
var result = new List <TaggedWord>();
var beforeTaggedWord = new TaggedWord("", "");
foreach (var taggedWord in sentence)
{
if (this.tokenizer.BeforeVerbs.Contains(taggedWord.word()) ||
(this.tokenizer.AfterVerbs.Contains(beforeTaggedWord.word()) &&
this.tokenizer.Verbs.Contains(taggedWord.word())))
{
beforeTaggedWord.setWord(taggedWord.word() + " " + taggedWord.word());
if (result.Count == 0)
{
result.Add(beforeTaggedWord);
}
}
else
{
result.Add(taggedWord);
beforeTaggedWord = taggedWord;
}
}
result.Reverse();
return(result);
}
/// <summary>
/// Join verb parts like Dadedgan corpus.
/// Input:
/// دیده/ADJ_INO
/// شد/V_PA
/// Iutput:
/// دیده شد/V_PA
/// </summary>
/// <param name="sentence">List of TaggedWord object </param>
/// <returns>List of TaggedWord</returns>
public static List<TaggedWord> JoinVerbParts(List<TaggedWord> sentence)
{
sentence.Reverse();
var result = new List<TaggedWord>();
var beforeTaggedWord = new TaggedWord("", "");
foreach (var taggedWord in sentence)
{
if (PeykareReader.tokenizer.BeforeVerbs.Contains(taggedWord.word()) ||
(PeykareReader.tokenizer.AfterVerbs.Contains(beforeTaggedWord.word()) &&
PeykareReader.tokenizer.Verbs.Contains(taggedWord.word())))
{
beforeTaggedWord.setWord(taggedWord.word() + " " + beforeTaggedWord.word());
if (result.Count == 0)
result.Add(beforeTaggedWord);
}
else
{
result.Add(taggedWord);
beforeTaggedWord = taggedWord;
}
}
result.Reverse();
return result;
}
public List <TaggedWord> BatchTag(List <string> sentence)
{
string[] sen = new string[sentence.Count];
for (int i = 0; i < sentence.Count; i++)
{
sen[i] = sentence[i].Replace(" ", "_");
}
List newSent = Sentence.toWordList(sen);
ArrayList taggedSentence = this._tagger.tagSentence(newSent);
var taggedSen = new List <TaggedWord>();
for (int i = 0; i < taggedSentence.size(); i++)
{
TaggedWord tw = (TaggedWord)taggedSentence.get(i);
tw.setWord(sentence[i]);
taggedSen.Add(tw);
}
return(taggedSen);
}
