Semantic Integration For Food Sales Analysis

Information systems interoperability is a complex topic that needs often to be addressed in a specific way. This paper presents an approach for interoperability between point-of-sales systems in the food sales domain and food composition database systems (food nutritional composition). The aim of this proposal is the creation of an integration mechanism able to associate food composition data from public health authorities to the food sold in point-of-sales systems. Assuming that food items identifications do not follow any identification standard, lexical, syntactical, semantic and classification techniques were introduced. Our intention is to demonstrate the feasibility of the automatic mapping process and the value raised by such an integration process.