• Predict the next word for any given phrase
  
• Demonstrate NLP modeling using R and Shiny
  
• Provide an interactive web app accessible online
  
• Inspired by predictive text systems like keyboards and chatbots

• Text cleaning and tokenization of English corpus
  
• Created n-gram frequency tables (uni-, bi-, tri-gram)
  
• Implemented back-off logic: trigram → bigram → unigram
  
• Simple prediction model for real-time response

• Built using R Shiny framework
  
• Input: user phrase → Output: predicted next word
  
• Lightweight and responsive UI
  
• Hosted on shinyapps.io

Live App:
https://nikhilchandra4321.shinyapps.io/text_prediction_app/

• Fast, minimal, and intuitive
  
• Example predictions:
  • “data” → science
    
  • “artificial” → intelligence
    
  • “machine” → learning
    
• Real-time feedback using reactive components

• Add top-3 suggestions
  
• Integrate deep learning models (LSTM / transformer)
  
• Train on larger text corpora
  
• Optimize performance for mobile devices