Data Science Capstone

For more detailed reference and source code, please visit: https://github.com/TeddyTiome/

• Word Database
• Pharse Process and Prediction
• Shiny Application

For a much more esay way to process the natural language, I have tried to serach the database for N-grams and dirty words.

Followed by these websites underlying, I have contributed 1 to 3 grams database as well as bad words.

• http://www.ngrams.info/
• http://storage.googleapis.com/books/ngrams/books/datasetsv2.html
• https://gist.github.com/jamiew/1112488
• Other pages on GitHub and searched by Google

Based on these database, the work is much easier, and the procedures include:

• Clean typed text with Non-alphabet characters.
• Filter bad words and set defualt to tokened tags.
• Get words in phrase and predict the next with most possibility.

Sorry that my shiny application is simple and concise.
So does the instructions:

• Type pharse, sentence, word without punctuation.
• Then the would do the prediction.
• A word would appear in the last text box.
  
  (I would say lazy-made and crude)