Recent years have seen growth of AI for IT Operations (AIOps) as an exciting new interdisciplinary research area. Many of the Knowledge Discovery and Data Mining technologies may play a critical role in AIOps. The purpose of this proposed interdisciplinary workshop is to bring together participants working on topics related to this new area from academia, industry, and government to provide a forum for them to discuss many frontier topics of knowledge discovery and data mining in the application context of AIOps, review the current state of the art, identify important challenges and opportunities for future research, and facilitate the creation of a new community in the intersection of AIOps and big data.

IT operations involve a set of people and management processes associated with IT service management to deliver the right set of services at the right quality and at competitive costs for customers. The operations and processes generate a large amount of data. For example, systems and applications regularly generate log data about usage patterns, activities and operations during running. Meanwhile metrics data is generated from performance monitoring systems which normally is time series data and provides an overall picture on the status of the IT environment. Anomalies may be observed from the log data and metrics data. Furthermore such anomalies may develop into significant events and incidents which need human attention. Tickets are then generated to keep track of human operations in the environment, such as problems encountered, analysis of root causes and solutions used to resolve the problems. Various machine learning and deep learning based methods have been proposed to handle aforementioned individual types of data. It is still extremely challenging to handle such large volumes of data at run time and identify correlations between them to automatically perform functions such as predicting potential outages and quickly allocating root causes when severe problems happen in an IT environment. Our focus on data mining and knowledge discovery is motivated by both the availability of large amounts of heterogeneous data sets in this domain and the great opportunities to use machine learning techniques to turn the data into knowledge and insights for optimizing IT operations. In addition, community members in the IT industry are leveraging user forums (e.g., Stack Overflow and Ask Ubuntu) to exchange insights and answers to IT issues and questions. The forums with millions of registered users and question-answer pairs provide tremendous value to the members who need help from experts on IT issues as well as researchers who want to turn content to actionable information for automation. This opens up many interesting research opportunities for natural language processing with noisy text data and machine learning involving humans in the loop. More recently, with the popularity of open source software (OSS) products and the trend of shift-left, it has become critical to help developers to find the most appropriate OSS products, automatically identify issues in the development cycle as early as possible (e.g., security and operational risks) and provide mitigation suggestions. Addressing these challenges requires intelligent algorithms for decision support as well as knowledge representation and inference in specific problem domains. All these open up many interesting new research opportunities for the BigData community.

Important Dates

  • Oct 8, 2022: Due date for paper submission
  • Nov 8, 2022: Notification of paper acceptance to authors
  • Nov 27, 2022: Camera-ready of accepted papers
  • Dec 17-20, 2022: Workshops