Projects

Tweet Spam Detector

The task of engaging customers across the globe for a multinational business can be enormous for a group of individuals to manage alone. This complexity has seen many companies turned to automated ways of engaging customers through the use of bots.

A study estimated that over 15% of Twitter users are automated bots.

For businesses that automate their interactions with their customers, they must do so with intelligent bots that can distinguish between what’s damaging and what’s not.

In this project, I built an LSTM model that is capable of analyzing the content of a tweet and categorize them into labels with 97% accuracy.

See project paper.