ANSI – Identifying consumer opinions on product safety and quality
The project designed ANSI, an AI-based system analysing multilingual online consumer opinions to identify concerns about product safety and quality, including dual quality, and recommend products for laboratory testing.
Lead Investigator: Łukasz Kobyliński
- Grant Number
- INFOSTRATEG-III/0005/2021
- Start Date
- 1 August 2022
- End Date
- 29 August 2025
Project factsheet
| English name: | Identifying consumer opinions on product safety and quality, including “dual quality”, using the OPIIPI system |
| Polish name: | Identyfikowanie opinii konsumentów na temat bezpieczeństwa produktów i ich jakości, z uwzględnieniem podwójnej jakości „dual quality” przy użyciu systemu OPIIPI |
| Project type: | A National Centre for Research and Development (NCBR) grant in the third call of the strategic research and development programme “Advanced information, telecommunication and mechatronic technologies” – INFOSTRATEG (commissioned projects) |
| Agreement number: | INFOSTRATEG-III/0005/2021 |
| Duration: | 1 August 2022 – 29 August 2025 |
| IPI PAN investigator: | Łukasz Kobyliński |
Consortium
- National Information Processing Institute – National Research Institute (OPI PIB), Warsaw
- Institute of Computer Science, Polish Academy of Sciences, Warsaw
The research topic of the call was submitted by the Office of Competition and Consumer Protection (UOKiK).
Project description
Goods sold on the Polish market in identical or similar packaging sometimes differ in composition or characteristics from those sold in other countries of the European Union. Such differences in product quality between markets, known as “dual quality”, conflict with consumer protection rules. The INFOSTRATEG III call, whose topic was proposed by UOKiK, concerned the use of artificial intelligence to identify consumer opinions on product safety and quality, including dual quality. The expected outcome was an IT tool analysing web resources – websites, marketplaces, blogs, forums and pages of consumer organisations – for information about product quality, whose results would guide laboratory testing of products and help monitor the situation in Poland and the EU.
ANSI, the system designed in the project by experts of OPI PIB and researchers of IPI PAN, relies on artificial intelligence methods, including machine and deep learning. It identifies products on the basis of multinational preferences and multilingual online consumer opinions, and the analysed data are used to issue recommendations on whether products should be examined further in a laboratory. Preparing the data for the system involved, among others, manual annotation, identifying websites containing product reviews, linking products with opinions, and identifying duplicated and synthetically generated reviews.
The call provided for three project phases, with NCBR selections after phases I and II. The project passed the selection after phase I (December 2023) and was concluded after phase II (August 2025).
In November 2022, ANSI was ranked third in the Innovative Product 2022 category of the Polish Radio ranking LUBIĘ, BO POLSKIE 2022.