Project Overview:

The goal of this project was to build a predictive text model that can accurately predict the next word in a given sentence using Natural Language Processing (NLP) techniques.

Benefit:

The Word Prediction App can assist users by predicting the next word as they type, improving typing speed and reducing errors. It can be particularly useful for mobile apps and chatbots, enhancing user experience.

Data Preprocessing:

• Cleaned and tokenized large text datasets (Twitter, news, and blogs).
• Applied profanity filters and removed unwanted words.

Model:

• Built an n-gram Markov Chain model with Kneser-Ney smoothing.
• Tuned the model to improve prediction accuracy.

Shiny App:

• Developed a Shiny app to interact with the model in real-time, offering word predictions as users type.

App Overview:

The Word Prediction App allows users to type sentences and get predictions for the next word using an intuitive interface.

https://szykora17.shinyapps.io/My_word_predictor/

Instructions:

1. Type a sentence in the text box.
2. The app will display the predicted next word.

How it works:

The app analyzes the input text and uses pre-trained n-gram models to predict the most likely next word. Based on user input, the app dynamically updates the predictions.

Thank you for your attention!

If you have any questions or would like to explore the app further, feel free to contact me.

Goodbye and good luck with your next word prediction endeavors!