return to curriculum vitae

Dagobert Soergel
College of Information Studies, University of Maryland

Improving access to food and nutrition data. 2. A language for the description of foods in databases

[Full Text PDF]

D Soergel, A. McCann, J.A.T. Pennington, J.H. Holden, E.C. Smith, and R.C. Wiley.
To be submitted to the Journal of Food Consumption Data.

This paper presents a highly flexible and powerful language for the description of foods in food composition and food use databases and for the precise specification of recipes and food standards. The language is the conceptual foundation for a proposed master database of food descriptions which would offer vastly improved capabilities for retrieval and combination of data from food composition and food use databases. The language incorporates principles from classification theory, database theory, and artificial intelligence. It uses the
entity-relationship approach to structuring food data in a way that mirrors the structure of the foods themselves. The entity types and relationship types needed to build food descriptions are detailed in the conceptual schema, and the most important ones are arrayed in a frame to guide indexers and searchers. The data structure uses hierarchical inheritance to avoid redundancy in storage. This paper also discusses food names and codes and their relationship to food description. It defines explicit names or codes that carry their own explanation and brief nonexplicit names or codes that merely refer to an explanation stored in one's memory or a database. In this context the paper discusses levels of specificity of the food products named, from product type to individual sample analyzed or portion consumed.
top of page