a) Sentence Segmentation - Since the sentences have been separated 
       by a dollar sign, we will separate them and store them in a 
       different array
    b) Tokenization
    c) Removal of Stop Words

    a) TF-IDF (Term Frequency- Inverse Document Frequency)
       - Compute the TF-IDF score of unigram, bigram and trigram.
       - Make a feature matrix
       - Split the data into training and testing. 
    b) Lexicon Based Approach
       - Compare with Hindi SentiWordNet and find the polarity
          and the POS tag as well.  

    a) Machine Learning Algorithms
       - k-Nearest Neighbors
    
    b) Lexicon-Based Approach
       - Use Hindi SentiWordNet(HSWN) to compute the Sentiment Score
       the sentence.