Machine Learning and Social Media

Public social media platforms have become very popular avenues for many people to get news to share their thoughts, feelings, and worldviews. In turn, social media data feeds can provide invaluable insights and strong signals of emerging problems. For example, leveraging powerful machine learning tools and social media feeds, we can detect when a social media account is involved with spreading misinformation, fake news, and hate speech, or in engaging in cyberbullying or malicious “trolling.” Additionally, we can predict when a user might have indications of mental health decline such as depression. Finally, we can detect when accounts on social media are being misused or abused for malicious purposes such as spamming, participating in cyberattacks, or proliferating malicious software. In this class, we will work with real world social media datasets (Twitter, Reddit, etc.) and applied machine learning techniques to develop models that indicate when a problem is under the way.