NSF POSE: Phase II: OpenAD: An Integrated Open-Source Ecosystem for Anomaly Detection

University Of Illinois At Chicago, Chicago IL

Investigators

Abstract

In the digital age, the ability to identify anomalies quickly and accurately within massive datasets is critical for ensuring the safety, security, and efficiency of numerous sectors, including healthcare, national security, and finance. Anomalies, which represent deviations from the norm, can indicate critical issues such as potential security breaches, health crises, or financial fraud. The open tools and systems for anomaly detection (AD) are often fragmented, complex, and not easily accessible. This project aims to revolutionize this landscape by developing an integrated, open-source ecosystem that simplifies AD. By making advanced detection tools widely available and user-friendly, the OpenAD ecosystem will empower researchers, businesses, and public institutions to leverage the full potential of AD. This project seeks to unify existing AD systems into a comprehensive ecosystem that supports diverse data types and application domains. This ecosystem will integrate a wide range of open-source AD tools, including those for tabular, graph, and time-series data, and provide a platform for seamless integration of model evaluation, automation, and acceleration. The project's goals include developing a standardized development environment, enhancing tools for automating and accelerating detection processes, and establishing a governance structure to guide community contributions and project evolution. By leveraging the collective expertise of a diverse community of developers, researchers, and users, OpenAD aims to overcome current limitations in AD. The resulting ecosystem will not only facilitate more effective and efficient detection of anomalies across various domains but also foster innovation and collaboration within the scientific community. Through OpenAD, AD tools will be more accessible, powerful, and capable of addressing the complex challenges of the increasingly data-driven world. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.