Printer Friendly

Bayesian Belief Networks offer advantages when assessing microbial safety.

The assessment of the microbial safety of foods should involve all available information and data relevant to a particular hazard. This information is almost always uncertain. Because biological systems are inherently variable, the best approach for undertaking this assessment involves probability-based modeling.

The goal of scientists at the Institute of Food Research is to develop tools for quantitatively assessing microbial risks. These tools should be easily accessible and user-friendly. Assessments must meet the standards maintained by international agencies such as Codex.

One interesting approach to assessing microbial safety is to use Bayesian Belief Networks (BBNs). BBNs are at the cutting edge of expert system technology. Unlike the traditional rule-based approach to expert systems, they are able to replicate the essential features of plausible reasoning--reasoning under conditions of uncertainty--in a consistent, efficient and mathematically sound manner. These networks are used to model complex beliefs and uncertainties.

BBNs are able to retract a belief in a particular case when the basis of that belief is explained away by new evidence. Since most real-life problems involve inherently uncertain relationships, BBN is a technology with huge potential for application across many domains. BBNs are based upon an expert system that exploits probability theory to provide a single framework for supporting multiple calculations and communications (including risks, costs and benefits). This approach also allows for transparent inspection and interrogation of each situation.

BBNs express influence, causation and dependency in a model domain. The networks offer an unbiased assessment of disparate sets of information. They facilitate the combination and quantification of connected uncertainties. Graphical representations of these networks promote improved inspections and communication.

The focus of scientists is to establish methods that support investigation into such areas as parametric and non-parametric learning from data, Bayesian inference in light of data which is not tractable to conventional statistical data analysis, Bayesian model discrimination, quantification of expert opinion, discrimination of variability and uncertainty from data. The safety issues of concern to scientists range from foodborne botulism--very rare but high-impact--to antimicrobial resistance in Campylobacter, which is an emerging issue. Scientists have developed expertise in chemical and microbiological risk assessments and are in the process developing Web-based tools.

Further information. Gary Barker, Institute of Food Research, Norwich Research Park, Colney, Norwich NR4 7UA, UK; phone: +44 1603 255000; fax: +44 1603 507723; email: gary.barker@bbsrc.ac.uk
COPYRIGHT 2005 Food Technology Intelligence, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2005, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Publication:Microbial Update International
Date:Aug 1, 2005
Words:388
Previous Article:Models predict the heat inactivation of L. monocytogenes in biofilms.
Next Article:Fighting L. monocytogenes with the bioprotective culture B-SF-43.


Related Articles
Identification and prevention of microbial risks in foods: the HACCP approach and priorities in Mexico.
Optimize risk assessment for ready-to-eat chilled products.
Quantitatively determine pathogen reduction.
FDA issues animal drug guidance.
Statistics software available from Univ of Washington.
Attributing illness to food.
Behavior of Listeria monocytogenes in pH-modified chicken salad during refrigerated storage.
Predictive microbiology targets poultry pathogens.
Innocent Targets.
Online training on mold.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters