A Data Driven Approach to Reducing Household Food Waste

Of all food wastes, that which is produced from the household is the most damaging in terms of environmental and economic impact. Many efforts have been made to quantify and analyse the reasons for and problems associated with household food waste generation which has led to the development of both technical solutions and behavioural interventions (including education and awareness) to try and reduce its generation. In this work a novel solution is proposed and developed which connects food providers and consumers, enabling more intelligent food planning, purchasing and consumption. A data driven Recipe Suggestion tool, supported by a Particle Swarm Optimisation (PSO) engine, is described for the first time. Recipes and associated ingredients are suggested for users which consider their preferences, remaining food items already held at home, expiry dates and minimum pack sizes. The tool is applied to a simulated case study to demonstrate its applicability and potential to generate a range of useful waste metrics. Results of the application of the tool, in terms of optimization capabilities and computation time, show encouraging potential for platform integration. The suitability of the tool to be incorporated into modern e-commerce systems is discussed.