The use of Kohonen's artificial neural networks for analyzing the results of HACCP system declarative survey

Abstract

The aim of this study was to apply the Kohonen's artificial neural networks (ANNs) for analyzing descriptive information on the HACCP system declared by surveyed food business operators. 66 businesses responded to the survey and completed the questionnaire regarding their own HACCP system. ANNs turned out to be a useful statistical tool for the analysis the results of the survey. It allowed the identification of similar opinions grouping them in clusters. Satisfactory were declarations of using more than one of HACCP principles verification method by the majority of surveyed businesses. However, dissatisfactory was often shown misunderstanding of GHPs and CPs as part of the HACCP system. HACCP principles were declared to be implemented mostly with the help of external consultants, which in terms of errors in the meaning of food safety management system indicated a constant need for training. It was impossible to separate the common opinion and build cluster applied to difficulties and benefits of the implementation of HACCP principles, which confirmed that each of the surveyed businesses represented a unique view on the functioning of their own system. This declarative study gave some overview on the knowledge on the HACCP system in the surveyed businesses, however it cannot be interpreted as a reflection of the true state of the system.