Defra - Department for Environment, Food and Rural Affairs.

Science Search

Science and Research Projects

Return to Science Search homepage   Return to Project List

Data mining for toxic hazard analysis (DTOX). - FS1719

The project will evaluate the utility of data mining techniques for predicting toxic hazards. A database of toxicological information will be produced, and data mining techniques will be used to generate prediction rules for a well-defined area of food chemical hazard assessment.
1. To determine by literature survey the chemical, biochemical and biological properties of compounds that are considered to be the most important for the prediction of toxic hazards. 2. To determine which of these properties are available in (or can be computed from) existing data, and thus to determine the areas of food chemical hazard assessment for which data mining methods are most likely to be useful. 3. To build a database orf the identified chemical and toxicological properties, for purposes of data mining. 4. To understand the informational properties of the available data, and using this understanding to convert the data into an appropriate form for data mining. 5. To use data mining to generate prototype rules predicting properties in the chosen area of hazard assessment. 6. To evaluate the performance of the prototype rules and to make recommendations on the feasibility and likely cost of a project to generate new rules for practical use by MAFF in food chemical hazard assessment.
Time-Scale and Cost
From: 1995

To: 1996

Cost: £99,598
Contractor / Funded Organisations
Integral Solutions Limited