Decision support for telecommunications and information society: Introduction

Abstract

Hunger remains a largely hidden social problem in many developed nations. The not-for-profit food rescue organizations, aid in alleviating hunger, by rescuing the surplus food from different food providers and re-distributing to people in need. However, surplus food donation is a random process which varies with regard to quantity, time and place. Understanding the dynamics of food recovery and forecasting food donations using historical information has significant importance in inventory management and redistribution, particularly in reducing operational costs and achieving a sustainable and equitable distribution of inventory incorporating uncertainties in supply. This paper uses different modelling techniques including multiple linear regression, structural equation modelling and neural networks to explore the patterns and dynamics of food donation and distribution process for one of the largest food rescue organization in Australia. A set of significant indicators has been identified to describe the current food donation process, to predict daily average food donated by different food providers and also to anticipate the potential donation from a new donor which may appear in the network in the future. Results suggest that structural equation modelling and neural networks provide improved demand estimation when compared to conventional multiple linear regression. We also discuss the usefulness of these models in sustainable and equitable management of food recovery and redistribution.