Typing on mobile devices is slow and error-prone.
This project builds a next-word prediction model similar to SwiftKey, using real-world English text data from blogs, news, and Twitter.

The goal is to improve typing speed by predicting the most likely next word based on previous input.

• Blogs – long-form writing
• News – formal language
• Twitter – short, informal text

Data was cleaned by: - Converting text to lowercase - Removing punctuation and numbers - Removing extra whitespace

• Built N-gram language models (1-gram, 2-gram, 3-gram)
• Calculated word probabilities
• Used back-off strategy when higher-order n-grams are unavailable

The model predicts the most probable next word for a given phrase.

• User enters a phrase in a text box
• Clicks Submit
• App returns a predicted next word

The app is deployed on shinyapps.io and responds in real time.

• Model demonstrates practical text prediction
• App provides a simple and intuitive interface

Future improvements: - Larger training data - Better smoothing techniques - Faster prediction response