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.
The dataset contains text from three sources:
These datasets contain millions of words used to build prediction patterns.
The algorithm uses n-gram language models.
Steps: - Clean and preprocess text - Tokenize words - Generate word frequency tables - Predict the most probable next word
The application was built using Shiny.
Features: - User enters a phrase - Clicks predict - The app returns the next word suggestion
App Link:
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