Studying the Presence of Genetically Modified Variants in Organic Oilseed Rape by using Relational Data Mining

The production of genetically-modified (GM) crops has increased rapidly over the last 10 years. The possibility of GM crops mixing with conventional or organic crops is becoming a problem and estimating the adventitious presence of GM seeds into conventional crop harvests presents a challenge. In this study we used outputs from a previously developed computer model for gene flow between GM and conventional oilseed rape to construct relational classification trees that predict the adventitious presence of GM seeds in the central field of a large-risk field pattern as a function of cultivation practices. Unlike propositional data mining methods, relational methods (relational classification trees) enable us to examine the relations among fields, for example, the influence of the neighbouring fields on the adventitious presence of GM seeds in a given field. For that purpose we used the relational data mining sys-