Words Correction of Sentence

Hi, i want to implement a post-processing for sentence correction from SVM Machine Learning.

I did know Machine Learning, but not in Natural Language Processing or Text Mining.

From Machine Learning, i got several words per sentence from handwritten text recognition.

But, there are several missclassification words that i did want to correction. The correction i need is :

Read the whole of sentence  
If there are double words besides or there are wrong grammar (Indonesian - SPOK) Then
      Read the whole of sentence for trace besides words
      Check the database of sentence with two words besides of center word AND get that center word
      Replace the center word of double words orwrong grammar with the center word
End if

Example :

Input

sentence = {'i', 'love', 'after', 'mother', 'she', 'me', 'me'};

Output

repalace_words = {'i', 'love', 'mother', 'after', 'she', 'kiss', 'me'};

[Angga Lisdiyanto]

Please help help help...