Disappearing Links in Social Media
Brief Description of Project:
The creation and disappearance of friendships on online social networks cause the evolution of social networks. Recently, researchers are beginning to acknowledge the importance of understanding broken ties online (e.g., unfollowing, blocking, etc.). Given they can also be used for better performance on computational tasks such as social recommendations, it is valuable to be able to predict them. This project will focus on the development of a deep learning approach (more specifically graph neural network) to predict which links will be broken in the near future, while also predicting the delay time for when these events will happen.
Programming experience (preferably in Python) and a strong interest in deep learning.
Nature of Supervision:
The student will work closely with the PI and have assistance from graduate students in the lab.
A Brief Research Plan (period is for 10 weeks):
Weeks 1-2: Project overview and background research
Weeks 3-8: Data curation and project implementation
Weeks 9-10: Project write-up
Number of Open Slots: 1
Name: Tyler Derr
Department: Electrical Engineering and Computer Science