Advertisement over Time: Interactive Machine Learning for Image Data Analysis
- Project Group:
DFG Individual Research Grant
This joint project with the research group of Prof. Florian Stahl at the University of Mannheim strives to advance research in both marketing and interactive machine learning. Hereby, we see a large potential for synergies, that we want to realize in this project.
Marketing researchers from Mannheim have acquired access to a data-set consisting of hundreds of thousands of scanned print advertisements (with a projected expansion to well over a million), ranging from the early 19th century to today. Their challenge lies within analyzing this data-set, to identify structural changes in what and how products are advertised. Further they want to investigate interdependencies with political and societal developments in this time frame. Our research group wants to leverage this undertaking as a context for interactive machine learning systems. Specifically, we want to design and develop systems that support the marketing researchers in labeling additional meta-data and features to augment their data-set. Another tool could support the analysis of such vast data-sets. Due to tight cooperation with our partners from Mannheim we are able to not only gather requirements for such tools from experts, but also to deploy to, and evaluate them with actual domain experts.
Project Related Publications
- Towards an Integrative Theoretical Framework of Interactive Machine Learning Systems. Meza Martínez, M. A.; Nadj, M.; Maedche, A. 2019. ECIS 2019 proceedings . 27th European Conference on Information Systems (ECIS), Stockholm & Uppsala, Sweden, June 8-14, 2019. Research Papers, Paper: 172, AISeL, Stockholm, Sweden
- Power to the Oracle? Design Principles for Interactive Labeling Systems in Machine Learning. Nadj, Nadj, M.; Knaeble, M.; Li, M. X.; Maedche, A. 2020. Künstliche Intelligenz, 1–12. doi:10.1007/s13218-020-00634-1
- Oracle or Teacher? A Systematic Overview of Research on Interactive Labeling for Machine Learning. Knaeble, M.; Nadj, M.; Maedche, A. 2020. 15. Internationale Tagung Wirtschaftsinformatik (WI 2020), Potsdam, 9 - 11 März 2020, 2–16, GITO Verlag. doi:10.30844/wi_2020_a1-knaeble
- Towards the Design of an Interactive Machine Learning System for Qualitative Coding. Rietz, T.; Maedche, A. 2020. ICIS 2020 – Making Digital Inclusive: Blending the Local and the Global, December 13-16, 2020
- Cody : An Interactive Machine Learning System for Qualitative Coding. Rietz, T.; Toreini, P.; Maedche, A. 2020. ACM Symposium on User Interface Software and Technology (UIST 2020), online, October 20–23, 2020. doi:10.1145/3379350.3416195
- Cody: An AI-Based System to Semi-Automate Coding for Qualitative Research. Rietz, T.; Maedche, A. 2021. Proceedings of the ACM Conference on Human Factors in Computing Systems (CHI 2021), Association for Computing Machinery (ACM). doi:10.1145/3411764.3445591