Sub-deliverable 2

Description: Other than Hidden Markov Model, I also did couple of experiments using Tanaka Corpus file.

Experiment 1: In this experiment, I am iterating through each and every character in the corpus file. When any special character is found, it has been stored as "EOW" (End Of Word). Each word is of length 3 and is stored as branches in binary search tree. Each node in binary search tree represents the word and total number of occurrences of that word. When user inputs the string, this program looks into the binary search tree to find the first three most relevant words with highest number of occurrences that are matching with the user input. Hence the aim of this experiment is to have suggestions of possible next character for the given user input.

Experiment 2: In this experiment, I am iterating through each and every character in the corpus file with string length of three. Along with this string, count is also stored. If the string ends with any special character then count is incremented else count is decremented. The string with highest positive count indicates boundary of the word. Hence the aim of this experiment to detect the end of the word.

Experiment 3: In this experiment, I am creating dictionary of japanese words using the Tanaka corpus file. I am iterating through each and every character in the corpus file unless any special character is been detected. Assuming special character marks end of word, program will store that word and its total number of occurrences will be updated.