Background

AI is applied across the entire retail value chain: demand and sales forecasting, warehouse and inventory logistics, replenishment, assortment and pricing optimization, personalization and customer service. The relevant literature is fragmented across information systems, operations research and computer science venues, and most existing reviews focus on a single application area or exclusively on either online or stationary retail.

A consolidated, up-to-date taxonomy structured along a recognized retail information-systems architecture is missing for the most recent wave of developments (foundation models, large language models, agentic systems).

Object of investigation

The goal of this thesis is to build a taxonomy of AI systems in retail following recent literature and industry trends. The taxonomy may be organized along the process and data structure of the Handels-H reference model (or similar reference models) and should classify each application along sensible dimensions to provide a structured overview of application areas for AI in retail organizations.

Possible research questions

• Which AI application areas in retail are documented in the recent literature, and how do they map onto the retail value chain?
• Which method families dominate each application area, and how mature is the available evidence?
• Where are the white spots that future research should address?

Possible results

• A taxonomy of AI systems in retail.
• A classified literature corpus and a derived research agenda.

Starting literature

• Brackmann, C., Hütsch, M. & Wulfert, T. (2023). Identifying Application Areas for Machine Learning in the Retail Sector. SN Computer Science, 4(5), Art. 426. DOI 10.1007/s42979-023-01888-w.
• Fildes, R., Ma, S. & Kolassa, S. (2022). Retail forecasting: Research and practice. International Journal of Forecasting, 38(4), 1283–1318. DOI 10.1016/j.ijforecast.2019.06.004.
• Becker, J. & Schütte, R. (2004). Handelsinformationssysteme, 2nd ed. Frankfurt a. M.: Redline Wirtschaft.