Forschung

Companies operate within an ever-evolving and changing environment, where their resilience hinges on their capacity for dynamic adaptation of business processes. Consequently, the proficient design and management of these processes acquire strategic significance. Amidst the digital transformation era, corporations are increasingly gravitating towards process-oriented information systems that document the execution of business process activities. To harness the full potential of such data, automated process analysis, known as process mining, serves as an efficacious tool. This form of analysis extends beyond merely reviewing past processes and includes forecasting future process outcomes (predictive business process monitoring).

Emerging in this context is the concept of Augmented Business Process Management Systems (ABPMs), representing an enhancement of process-oriented systems through the integration of reliable artificial intelligence. ABPMs aim to render the process lifecycle more adaptive, proactive, explainable, and sensitive to contextual nuances. The confluence of automation within the domain of business processes opens up novel avenues to identify and capitalize on process improvement opportunities. This synergy of productive process redesign, facilitated by artificial intelligence, and the automation of these processes using software robots (Robotic Process Automation or RPA), along with their coordination, is collectively termed hyperautomation.

Artificial Intelligence (AI)-based Decision Support Systems (DSS) are increasingly being utilized to assist in complex decision-making across various domains. AI encompasses a concept of data-driven problem-solving through mathematical algorithms, frequently linked to the realms of Machine Learning (ML) and Deep Learning (DL). Specifically, DL methodologies employ increasingly intricate computational logics to enhance performance outcomes. Concurrently, there is a discernible preference in practical applications for algorithms with reduced complexity, thereby enhancing comprehensibility. The research field of Explainable AI (XAI) concentrates on demystifying complex AI algorithms, offering insights into the underlying decision-making logic of the models employed. As the complexity escalates, there is often a corresponding increase in the volume of training data required to adequately train these models. In many smaller enterprises, where such extensive data may not be readily available, the use of pre-trained models through transfer learning is common. These models are then fine-tuned to specific needs. Our chair particularly focuses on the application of AI within the context of business information systems and hyperautomation, along with specialized focus areas in XAI and transfer learning.