• Natural Language Processing App
  • All code implemented using R
  • Hosted at https://www.shinyapps.io \[\\\]
• Goal: Predict third word of tri-gram given two leading words \[\\\]
• Use Katz Back-Off Method for predictions
  • Provide list of word predictions along with probabilities

• Three (very large) starter files
  • en_US.blogs.txt
  • en_US.news.txt
  • en_US.twitter.txt
• Create a sample dataset using 5% each of the .txt files
  • Corpus created and cleaned using the quanteda package.
• Katz Back-Off Prediction Method
  • A generative n-gram language model that estimates the conditional probability of a word given its history in the n-gram of a training corpus.

Screenshot from https://roarmarketingconcepts.shinyapps.io/CapstoneProject-ShinyApp/

• Predicting Next Word Using Katz Backoff - Michael Szczepaniak \[\\\]
• Katz’s Backoff Model - Wikipedia article \[\\\]
• Developing Data Products in R Course - Johns Hopkins University on Coursera