Data Science Capstone Project

Next Word Prediction

Data Exploration

The project uses three English text datasets:

  • Blogs
  • News
  • Twitter

The datasets were explored to understand text size, word frequency, and differences between sources.

Prediction Algorithm

A next-word prediction model was developed using n-gram based text analysis.

The model uses:

  • Bigram
  • Trigram
  • Four-gram

The most likely next word is selected from the observed word combinations.

Shiny Application

The Shiny app allows users to:

  • Enter a phrase
  • Click “Predict Next Word”
  • Get the predicted next word

Example:

I want to → see

The application is deployed on shinyapps.io.

Conclusion

The project successfully developed and deployed a next-word prediction application.

Key Outcomes:

  • Text data explored
  • N-gram prediction model created
  • Shiny application developed
  • Application successfully deployed

Shiny App:

https://gaganpandey921.shinyapps.io/DataScienceCapstone/