AIOPS 2023

International Workshop, co-organized with ICDM 2023

Shanghai, China, December 01, 2023

DOS TU Berlin

---

Modern IT systems, such as data centres, cloud computing environments, edge clouds, IoT, and embedded environments, are the key enablers of digital transformation. Their complexity in development and management puts an enormous burden on the developers and operators to efficiently instrument the source code and leads to a large volume of data subjected to analysis. This oftentimes makes the localization of the most important clues for problem resolution challenging, severely increasing the potential negative impact of the failures, (e.g., large economic losses or threatening human safety). Due to ever-increasing complexity, both developers and operators increasingly rely on tools from artificial intelligence for assistance during IT system development and operation.

Artificial Intelligence for IT Operations (AIOps) is an emerging field arising at the intersection between the research areas of data mining, machine learning, big data, streaming analytics, and the management of IT operations. The main goal is the analysis of the heterogeneous data exposed by the IT systems in the form of semi-structured text (e.g., system logs, source code), data with temporal and spatial characteristics (e.g., KPI metrics, component dependencies), streams of heterogenous types, and similar data, to support the objectives such as prevention of SLA violation, early anomaly detection and auto-remediation, energy-efficient system operation, providing optimal QoE to customers, predictive maintenance and many more. Recent advances in intelligent coding are opening up new possibilities for efficiently instrumenting source code while maintaining reliable computing characteristics. We envision that, with the advance of AIOps technologies, and incorporating the scaling, efficiency and security in their design, the IT industry will achieve significant progress and sustained exponential growth.

In continuation of the series of our prior successful workshops, we intend to organize the third workshop on artificial intelligence for software development and IT operations. We aim to gather researchers from academia and industry to present their experiences, results, and work in progress in this field. Of particular, but not limited, interest is the proposal of novel and adoption of existing data mining, machine learning and big data techniques for the emerging problems of auto-instrumentation, efficient utilization of open telemetry, techniques for software coding, testing on the fly, and building efficient visualizations. In this regard, our goal is to provide a venue for discussing the challenges in AIOps and create a community roadmap for a response to the most important challenges. To enable a direct and fruitful discussion, we aim for a selected number of publications through a rigorous peer-review procedure from renowned experts in the area. We strongly encourage the authors to adopt the principles of open science and reproducibility to boost research in this emerging area.

• Autonomous instrumentation
• Safe and reliable intelligent software coding
• Log analysis
• Anomaly detection
• Failure mode analysis
• Self-healing, self-correction and auto-remediation
• Benchmarking in AIOps
• Hardware and software failure prediction
• Root cause analysis
• Performance management
• Predictive and prescriptive maintenance
• Resiliency, reliability, and quality assurance
• IT system dependability
• Energy-efficient cloud operations
• Resource management
• Autonomous service provisioning
• Visual analytics and interactive machine learning
• Fault injection, verification testing and chaos engineering
• Use-cases, testbeds, evaluation scenarios

Important Dates

The focus of the workshop involves, but it is not limited to the following topics:

• Submission deadline: September 8th, 2023
• Notification deadline: September 22nd, 2023
• Workshop: one day between December 1-4, 2023 in Shanghai, China