runtime: shiny

title: "Next Word Prediction App" author: "doctahmanhattan" date: "07/27/2024" output: ioslides_presentation ---

Introduction

• Objective: To build a next word prediction model.
• Tools: R, Shiny, N-gram models.

Algorithm

• Based on n-gram models.
• Uses historical data from blogs, news, and Twitter.

App Description

• Text input for phrase.

• Predicts the next word.

• Deployed on shinyapps.io.

  App Code

  ``` library(shiny) library(tidyverse)

Define UI for application

ui <- fluidPage( titlePanel("Next Word Predictor"), sidebarLayout( sidebarPanel( textInput("text", "Enter a phrase:", ""), actionButton("predictButton", "Predict Next Word") ), mainPanel( textOutput("prediction") ) ) )

Placeholder prediction function (replace with your actual prediction logic)

predictnextword <- function(input_text) { # Implement your prediction algorithm here # For example, using n-gram model return("example") # Replace with actual predicted word }

Define server logic

server <- function(input, output) { observeEvent(input$predictButton, { output$prediction <- renderText({ req(input$text) predictedword <- predictnextword(input$text) paste("Next word prediction:", predictedword) }) }) }

Run the application

shinyApp(ui = ui, server = server) ```

Examples

• Phrase: "I love to"

• Prediction: "read"

• Phrase: "This is a"

• Prediction: "test"

Conclusion

• Successful prediction model.
• Future improvements: more data, better algorithms.