This website requires JavaScript.

Studying the Characteristics of AIOps Projects on GitHub

Roozbeh AghiliHeng LiFoutse Khomh
Dec 2022
Artificial Intelligence for IT Operations (AIOps) leverages AI approaches tohandle the massive data generated during the operations of software systems.Prior works have proposed various AIOps solutions to support different tasks insystem operations and maintenance (e.g., anomaly detection). In this work, weinvestigate open-source AIOps projects in-depth to understand thecharacteristics of AIOps in practice. We first carefully identify a set ofAIOps projects from GitHub and analyze their repository metrics (e.g., the usedprogramming languages). Then, we qualitatively study the projects to understandtheir input data, analysis techniques, and goals. Finally, we analyze thequality of these projects using different quality metrics, such as the numberof bugs. We also sample two sets of baseline projects from GitHub: a randomsample of machine learning projects, and a random sample of general purposeprojects. We compare different metrics of our identified AIOps projects withthese baselines. Our results show a recent and growing interest in AIOpssolutions. However, the quality metrics indicate that AIOps projects sufferfrom more issues than our baseline projects. We also pinpoint the most commonissues in AIOps approaches and discuss the possible solutions to overcome them.Our findings help practitioners and researchers understand the current state ofAIOps practices and sheds light to different ways to improve AIOps weakaspects. To the best of our knowledge, this work is the first to characterizeopen source AIOps projects.