Next-Word Prediction App

🔍 What is this project?

• A Shiny web app that predicts the next word given a phrase.
• Built with R using NLP techniques.
• Demonstrates the power of N-gram models and Shiny interfaces.

Use cases:

• Autocomplete typing
  
• Chatbots and virtual assistants
  
• Smart prediction in mobile devices
  

⚙️ N-gram Backoff Model

• Uses 1-gram, 2-gram, and 3-gram frequency tables.
• Cleans and tokenizes user input.
• Applies a backoff strategy:
  • Tries 3-gram first
  • Falls back to 2-gram
  • Then falls back to 1-gram

predict_next_word <- function(input_text) {
  # Example placeholder
  return("example")
}

🖥️ User Interaction Flow

• The user types a phrase in the text input field.
• The app processes the input and predicts the next word.
• The predicted word appears below the input.

Example Phrases and Predictions:

📚 Why R?

• 🧹 Data cleaning: stringr, textclean
• 🔢 Tokenization: tidytext
• ⚡ Speed: data.table for large n-gram datasets
• 📊 Visualization: ggplot2
• 🌐 Web app deployment: shiny, shinyapps.io