2026-06-30

Slide 1: Title

Next Word Prediction Application

Data Science Capstone Project

Your Name

Slide 2: Project Overview

Objective

  • Predict the next word from a user-entered phrase.
  • Use an n-gram language model.
  • Provide predictions through a Shiny application.

Data

  • English Blogs
  • English News
  • English Twitter

Slide 3: Prediction Algorithm

Method

  1. Convert text to lowercase.
  2. Remove punctuation.
  3. Search the trigram table.
  4. If no match, search the bigram table.
  5. If still no match, return the most frequent unigram.

Example

Input:

I love

Prediction:

you

Slide 4: Shiny Application

Features

  • Text input box
  • Predict button
  • Displays the predicted next word
  • Fast response using precomputed n-gram tables

Screenshot

(Insert a screenshot of your Shiny app here.)

Slide 5: Conclusion

Summary

  • Built a next-word prediction model.
  • Developed an interactive Shiny application.
  • Used a backoff strategy to improve prediction coverage.

Future Improvements

  • Larger training dataset
  • Four-gram or five-gram models
  • Better handling of unknown words