• Typing on mobile devices is often slow and error-prone.
• Users spend too much time correcting errors and retyping common phrases.

• Developed a predictive text algorithm using bigram modeling.
• Our Algorithm has the capability to predict the next two words based on user input.

• Use a subset of data from twitter, news, and blogs.
• Clean and processed the data to remove noise.
• Bigrams generated from subset of the combined datasets (twitter, news, and blogs)
• Predictions based on the most frequent word pairs in the text data.

• Improves typing speed and reduces errors.

Next Steps

• Deploy app at scale and gather user feedback.
• Refine the algorithm with more data.