This capstone project demonstrates a next-word prediction application developed using R and Shiny. It uses English text datasets from blogs, news, and Twitter to build bigram, trigram, and fourgram frequency models. The presentation covers data preprocessing, the prediction methodology, the interactive Shiny application, results, and future improvements.

This report presents an exploratory data analysis of the SwiftKey dataset, including US English blogs, news, and Twitter text files. It examines the basic structure and characteristics of the data, performs text analysis, and highlights important patterns in word usage. The findings will help guide the development of a predictive text algorithm and a Shiny application that suggests the next word based on user input.