SentenceParseNew : Parts Of Speech tagger

Takes a sentence as input and returns the POS tag for each word

Requirements :
1) function checker_first : user written
2) function checker_rest : user written
3) function most_prob : user written
4) TrainingDataSPNew
All these files are also available.

The program takes a string as input from the user, and returns the Parts Of Speech tag for every word in the sentence. It relies on the Training Data in TrainingDataSPNew, which is updated every time the sentence has been tagged (after the user confirms the tags, to prevent mistagging). It uses a Bigram model where the most probable tag for a word is calculated based on the tags present in the training data for that word, and also based on the tag of the word before the word in concern, in the input sentence. For example, as we know, a noun is usually preceded by an article, so if given a new noun (not in the training data), the program sees that the word preceding it was an article and so tags the word as a noun.
If any queries, errors or suggestions, do mail me at samyuktaramnath[at]gmail[dot]com, or at hale2bopp[at]gmail[dot]com.