Data Science Capstone Project : Predicting Next Word

The Shiny App is an R Shiny application exploring the area of text prediction, in part fulfillment of the requirements of the Coursera JHU Data Science Specialization Capstone Project.

The following objectives were considered and achieved in the design of this application:

• Mandatory requirement of predicting next word given an input n-gram
• Ease of use for general users
• Fast and lightweight

The following steps were performed in building the application:

• The Coursera dataset was cleaned (Whitespaces, non-alphabetic characters were removed, and all text are converted to lower case) {2,3,4}-grams were extracted and formed for our usage in the prediction algorithm.
• The best match (ordered by n-Gram length, followed by prevalence) is utilized in predicting the next word following the input n-Gram. “