Inspiration
Text messages are the contemporary equivalent of writing letters. The messages we send through various communication protocols are a reflection of who we are. Through modern analytical techniques, the tools exist to draw conclusions about the ways we message people and to provide predictions using machine learning to assist our messaging habits.
What it does
Provides analysis of a user's text messages or FB messages by grouping a user's messages with their friends into units of "conversations." This analysis includes a user's most commonly used words and sentiment values of their messages with their friends. Further, machine learning allows for the prediction of who a user wants to send a message to as they're writing it.
How we built it
Messages were formatted into CSV files which served as our data sets. These files were then further cleaned, and relevant data points were extracted. This data was parsed for frequency of word usage and sent to Microsoft Cognitive Services for analysis such as sentiment values of conversations. Microsoft Azure Machine Learning was used to process the data to determine key phrases which allows for prediction of who a user is sending a message to.
Log in or sign up for Devpost to join the conversation.