Novel Automated System for the Compliance Assessment of E-foods Merchandised in Germany

Sunday, February 14, 2016
Alexandra Krewinkel, Hamburg University of Applied Sciences, Hamburg, Germany
The online grocery grew fast and the demand of foods on the Internet continues to increase steadily. By now the offer of online food retailers ranges from specialities and niche products within small up to full-range suppliers in high professional shops with thousands of subpages. Grocery retailing on the Internet (e-foods) is subject to the same national and European food laws and regulations as conventionally merchandised foods. Consequently, the competent authorities responsible for the food control have to face new challenges. Currently food inspection has to manage huge efforts to monitor the compliance of e-foods by doing manually performed online searches in order to find potential infringements. To reduce these enormous efforts and to analyse big data it is unavoidable to develop IT-based strategies capable of managing these novel challenges. For this reason the Hamburg University of Applied Sciences has started the research project “Development of automated analysis procedures for identification and assessment of non-compliant products of the virtual food market” (AAPVL) in cooperation with the Federal Office of Consumer Protection and Food Safety (BVL). The aim of the project is firstly to develop a conceptual model enabling an automatic collection, classification, and extraction of information obtained from internet searches and secondly to develop a software prototype based on that model. The concept of monitoring e-food consists of a multistage data process applying various filter steps. The novel IT-based system for the compliance assessment of e-foods will be presented. Moreover, first results of case studies applying the novel concept for the compliance assessment of e-foods will be presented. Applying the software prototype a data set of 8,683 search results attributed to pharmacologically active substances synephrine and coffeine were reduced to 1,242 suspicions offers representing in total 219 product offers offered by 440 vendors. Thus, the novel software prototype is the technological bases for the development of an automated analysis tool for e-foods contributing to a higher level of food safety in online grocery.