Next Word Prediction App

Afreen Mujawar

Overview

This project builds a predictive text model similar to smartphone keyboards.

The application predicts the next word based on the phrase entered by the user.

Dataset

The dataset contains text from three sources:

  • Blogs
  • News articles
  • Twitter posts

These datasets contain millions of words used to build prediction patterns.

Prediction Algorithm

The algorithm uses n-gram language models.

Steps: - Clean and preprocess text - Tokenize words - Generate word frequency tables - Predict the most probable next word

Shiny App

The application was built using Shiny.

Features: - User enters a phrase - Clicks predict - The app returns the next word suggestion

App Link:

https://afreenmujawar1997.shinyapps.io/wordPredictionApp/

Conclusion

This project demonstrates how natural language processing techniques can power predictive typing systems.

Future improvements include: - Larger n-gram models - More accurate predictions - Multiple word suggestions