Automatic classification of takeaway food outlet cuisine type using machine (deep) learning

Tom R.P Bishop,Stephanie Von Hinke,Bruce Hollingsworth,Amelia A Lake,Heather Brown,Thomas Burgoine

doi:10.1016/j.mlwa.2021.100106

Abstract

Background and purpose:Researchers have not disaggregated neighbourhood exposure to takeaway (‘fast-’) food outlets by cuisine type sold, which would otherwise permit examination of differential impacts on diet, obesity and related disease. This is partly due to the substantial resource challenge of manual classification of unclassified takeaway outlets at scale. We describe the development of a new model to automatically classify takeaway food outlets, by 10 major cuisine types, based on business name alone. Material and methods:We used machine (deep) learning, and specifically a Long Short Term Memory variant of a Recurrent Neural Network, to develop a predictive model trained on labelled outlets (n=14,145), from an online takeaway food ordering platform. We validated the accuracy of predictions on unseen labelled outlets (n=4,000) from the same source. Results:Although accuracy of prediction varied by cuisine type, overall the model (or ‘classifier’) made a correct prediction approximately three out of four times. We demonstrated the potential of the classifier to public health researchers and for surveillance to support decision-making, through using it to characterise nearly 55,000 takeaway food outlets in England by cuisine type, for the first time. Conclusions:Although imperfect, we successfully developed a model to classify takeaway food outlets, by 10 major cuisine types, from business name alone, using innovative data science methods. We have made the model available for use elsewhere by others, including in other contexts and to characterise other types of food outlets, and for further development.

Highlights

Background and purposeResearchers have not disaggregated neighbourhood exposure to takeaway (‘fast-’) food outlets by cuisine type sold, which would otherwise permit examination of differential impacts on diet, obesity and related disease
While a growing number of studies have demonstrated an association of neighbourhood exposure to unhealthy takeaway food outlets with poor diet, greater body weight and odds of obesity (Burgoine et al, 2016; Burgoine, Forouhi, Griffin, Wareham, & Monsivais, 2014; Burgoine, Sarkar, Webster, & Monsivais, 2018), the evidence base remains equivocal (Black, Moon, & Baird, 2013; Fleischhacker, Evenson, Rodriguez, & Ammerman, 2011; Wilkins et al, 2019)
Our model had a cuisine type classification recall of 72% (‘‘good’’), with 72% precision (Table 2). This is to say, 72% of outlets had their cuisine type correctly predicted, and out of all predictions made by the model, 72% of those were correct

Summary

Introduction

Background and purposeResearchers have not disaggregated neighbourhood exposure to takeaway (‘fast-’) food outlets by cuisine type sold, which would otherwise permit examination of differential impacts on diet, obesity and related disease. While a growing number of studies have demonstrated an association of neighbourhood exposure to unhealthy takeaway food outlets with poor diet, greater body weight and odds of obesity (Burgoine et al, 2016; Burgoine, Forouhi, Griffin, Wareham, & Monsivais, 2014; Burgoine, Sarkar, Webster, & Monsivais, 2018), the evidence base remains equivocal (Black, Moon, & Baird, 2013; Fleischhacker, Evenson, Rodriguez, & Ammerman, 2011; Wilkins et al, 2019) In some instances, this may be the result of exposure misclassification i.e. incorrect specification of a causally relevant environmental exposure (Cummins, Clary, & Shareck, 2017), which serves to mask true associations and potentially biases any observed associations towards the null (Hutcheon, Chiolero, & Hanley, 2010). We use mid-2019 population estimates from the Office for National Statistics

Objectives

Methods

Results

Discussion

Conclusion