Publication Date

Spring 2019

Degree Type

Master's Project

Degree Name

Master of Science (MS)


Computer Science

First Advisor

Katerina Potika

Second Advisor

Christopher Pollett

Third Advisor

Rula Khayrallah


Online Social Networks, Network Alignment, Supervised learning, Unsupervised learning


Online Social Networks (OSN) have numerous applications and an ever growing user base. This has led to users being a part of multiple social networks at the same time. Identifying a similar user from one social network on another social network will give in- formation about a user’s behavior on different platforms. It further helps in community detection and link prediction tasks. The process of identifying or aligning users in multiple networks is called Network Alignment. More the information we have about the nodes / users better the results of Network Alignment. Unlike other related work in this field that use features like location, timestamp, bag of words, our proposed solution to the Network Alignment problem primarily uses information that is easily available which is the topology of the given network. We look to improve the alignment results by using more information on users like username and profile image features. Experiments are performed to compare the proposed solution in both unsupervised and supervised setting.