Network science and AI help us understand, model, and predict human behavior. Where network science focuses on the patterns of relations between entities such as people, ideas, organizations and so forth using graph and statistical methods. AI (including machine learning, computational linguistics, and large language models) is often used to focus on the content of the messages shared during interactions. Both methods have limitations. In this course we explore how they can be used together to overcome these limitations.

This course provides an introduction to network science and how network science is enabled by artificial intelligence (AI). Topics that will be covered include identification of key actors and groups, stance, network comparison, and network dynamics. AI will be used to generate synthetic network data, label groups, and identify missing links in networks. This course will provide an overview of how network science can be used to overcome limitations in AI systems and how AI can be used to overcome limitations in network data and support analysis. Much of the training will be hands-on and participants will be given data and technologies to analyze. The data provided will be organized in scenarios that the participants will analyze and produce insights related to as they use the AI enabled network science methods and tools provided.

Syllabus

https://www.cmu.edu/casos-center/events/courses/17-920-ai-enabled-network-science-syllabus1.pdf

Class format

Lecture and project-based

Home department

Software and Societal Systems

 Background required

No prerequisites

Faculty and instructors who have taught this course in the past

Kathleen M. Carley