This project builds a predictive model to suggest the next word in a user-input phrase. It uses NLP techniques trained on large text corpora to improve user text input efficiency.

We use an n-gram model (bi- and tri-gram) with a backoff strategy. If a trigram is not found, the model falls back to a bigram. If no match is found, a common word is returned.

• Input: Any partial English phrase
• Output: A predicted next word
• Interface: Built with R Shiny
• Hosted on: shinyapps.io

1. Enter a phrase in the text box.
2. Click “Predict” to get the next word.
3. Example: “How are” → “you”

• Simple and functional user interface
• Lightweight and responsive
• Useful for predictive text applications
• Ready for integration and scaling