This project presents a predictive text application that can suggest the next word based on the user input text.

The app is deployed as a Shiny application, which makes it available for access in modern web browsers.

Typing on devices like laptops can be slow as well as error prone. Chances of errors are higher in long messages or posts.

This app will provide suggestions to make typing quicker and less error prone as it is made by analyzing large text datasets from Twitter (or X), news, and blogs.

The model uses trigrams, bigrams, and unigrams.

This application predicts the next word based on the last two words, falls back to the last word, and finally uses the most frequent word. This ensures accurate predictions while staying fast.

The predictive text app suggests the next word based on what the user types. Users type a phrase and click the “Predict” button, and the app displays the most likely next word.

For example, typing “I love to” predicts “see”, “The stock market” predicts “and”, and “She went to” predicts “the”. This shows how the app can help users type faster and reduce errors by offering context-aware suggestions.

The app can be improved by using bigger datasets and advanced prediction methods.

Adding features like personalized suggestions, phrase prediction, and filtering inappropriate words would make typing faster, more accurate, and safer.