Pubblicazioni

Connecting entries from environmental and nutritional databases of food products is needed to identify sustainable food options. To overcome the hurdle for a successful data standardization, this paper aimed to explore the general structure of food composition and life cycle inventory (LCI) databases and to provide a semiautomatized approach on how to successfully interlink data of two example databases. The structure, the data availability and accessibility of food entries (FI) from the LCI database Agribalyse and selected food composition EuroFIR databases were analyzed. Harmonized food names from LanguaL™ codes from food classification systems were gathered and validated manually in order to use as descriptors to tag database entries in an automated way. Both databases, EuroFIR and Agribalyse, provided sufficient amount of meta data to interlink FI with the standardization approach proposed. Information on food name, food specification, food processing and the type of production was used for data interlinkage purposes. Manual validation of data interlinkage showed that two out of a sample of 54 entries were found to have incorrectly assigned descriptors. Agreeing on common principles (e.g., use of a specific classification system, common database formats) and improving meta data availability would facilitate database interlinkage and improve both, accuracy and efficiency. Developing solutions to increase meta data availability and accessibility of FI in food databases should become a key area of research in order to transition into more reliable database connection systems.