Comparative research on food products has shown that almost every tenth product sold in the European Union is of double quality. Although according to EU law, goods should be of the same quality, unfortunately, they often have very different quality parameters. The one on the Polish market usually has them worse. Thus, Polish scientists have developed a solution to battle such practices in the form of artificial intelligence – ANSI system.
“Dual quality is a practice in which companies use different recipes, formulations or standards for items sold under the same brand name and with very similar looking packaging”. (safefoodadvocacy.eu)
Experts from National Information Processing Institute (OPI) together with scientists from the Institute of Computer Science of the Polish Academy of Sciences (IPI PAN) designed the ANSI IT system. The tool will operate on the basis of methods and techniques in the field of artificial intelligence, especially machine and deep learning. The basis of its operation will be the identification of products on the basis of multinational preferences and multilingual consumer opinions published on the Internet.
Work on the development and implementation of the ANSI system is to be completed by 2025 at the latest.
“Thanks to the use of the latest methods of natural language processing, we are able to identify products whose quality in Poland should be additionally tested in a laboratory, not by directly analyzing each physical product separately, but indirectly, on the basis of opinions about these products available on the Internet. This is a very interesting application of artificial intelligence methods, which in a positive sense uses the fact that so much of our life is now transferred to the virtual sphere.”, says Ph.D. Łukasz Kobyliński from the Linguistic Engineering Team of IPI PAN, head of tasks in the ANSI project.
The ANSI tool, after a thorough analysis of consumer opinions, will select and create recommendations for a given product suspected of dual quality to be sent for laboratory verification. It will also offer a platform that allows users to access the collected data and analysis results. The project will result in the creation of a high-quality database of products and opinions about them, along with a list of laboratories.