This project focuses on developing a Next Word Prediction App using Natural Language Processing (NLP) techniques in R.

The application is built using Shiny and the report is presented through R Markdown using a slide-based layout.

• The app predicts the next likely word as a user types text.
• Uses N-Gram language modeling and Stupid Backoff algorithm.
• Built using text data from Blogs, News, and Twitter.
• Deployed online as an interactive Shiny application.
• Helps enhance user typing experience similar to mobile keyboards.

Click the Below Link for the Application.

https://ndeekshith.shinyapps.io/dscapstone/

Click the link below for compiled project files on gitub repo.

https://github.com/ndeekshith/Data-Science-Capstone-Peer-Graded-Assignment

The data used for this project comes from the HC Corpora dataset, which includes text from:

• Blogs
• News articles
• Twitter posts

The original dataset is very large (over 500MB). To optimize performance:

• A small random sample (2%) was taken from each source
• Profanity and unwanted characters were removed
• Data was tokenized to build N-gram language models

library(quanteda)
summary(corpus_data)