ÖIAT Research

The Austrian Institute for Applied Telecommunications (ÖIAT) has been promoting the competent, safe and responsible use of digital media for more than 20 years. Together with our partners, we conduct research on current topics related to the digital world. On this page we provide an overview of past and current research projects.

eMarketShield


go to emarketshield.at
Coordination
go to ages.at
Partner
go to www.sozialministerium.at
go to fh-ooe.at
go to liqua.net
go to oiat.at
go to x-net.at
go to xylem-technologies.com
Funding
go to kiras.at
go to ffg.at
go to bmf.gv.at
Contact

Julia Krickl
Co-Head of Research & Innovation
krickl@oiat.at
+43-1-595 2112-28

2024/102026/09
Consumer Protection Food Safety

The distribution of dangerous or unauthorized products as well as misleading, false promises pose safety risks for consumers in Austria – and are currently inadequately addressed on the internet. The eMarketShield project aims to enhance customer safety in online commerce, with a specific focus on dietary supplements and goods subject to the Food Safety and Consumer Protection Act (LMSVG). The results are intended to contribute to market surveillance for all product categories sold online.


The eMarketShield project aims to develop and evaluate risk-based control approaches tailored to the requirements of official online commerce inspections using data-driven decision-making. Alongside crawling, scraping, and API access, current methods of Artificial Intelligence (AI) such as Large Language Models (LLMs) will be employed to identify non-compliant products, detect trends early, and efficiently utilize limited resources through automatic scoring. Additionally, special attention is given to promoting Austria's digital sovereignty by developing and implementing these technologies and control approaches within a national context.

The project builds on proven approaches of risk-based control in traditional retail. For selected use cases, own approaches are implemented during data collection through crawling, scraping, and accessing existing data and public APIs, which are then compared with existing solutions. One focus is the use of LLMs, which can revolutionize the processing of unstructured web data. Data utilization is based on a Common Data Model conceived within the project, ensuring a uniform structure and future exchange. To analyze trends in the data, algorithms such as Top2Vec and LLMs are used. Alternative approaches to resource-intensive sampling and mystery shopping, such as prevention through information to consumers or platforms/sellers, are evaluated through triage with regard to improving consumer protection in a prototype web application.

The results of this project provide a solid, data-driven basis for planning tool support for market surveillance in online commerce in Austria. They not only identify potential weaknesses in internet market surveillance but also provide solutions to optimize and increase process efficiency. The developed technological components and methods thus form a solid basis for informed considerations regarding further development towards market readiness. These results are crucial for effectively managing official control in online commerce and minimizing potential risks, both technically and legally and ethically.

eMarketShield is funded by the KIRAS security research programme of the Federal Ministry of Finance.