Browse by author
Lookup NU author(s): Dr Heather Brown
This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).
Background and purpose Neighbourhood exposure to takeaway (‘fast’-) food outlets selling different cuisines may be differentially associated with diet, obesity and related disease, and contributing to population health inequalities. However research studies have not disaggregated takeaways by cuisine type. This is partly due to the substantial resource challenge of de novo 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=4000) 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.
Author(s): Bishop RPT, von Hinke S, Hollingsworth B, Lake AA, Brown H, Burgoine T
Publication type: Article
Publication status: Published
Journal: Machine Learning with Applications
Print publication date: 15/12/2021
Online publication date: 10/07/2021
Acceptance date: 05/07/2021
Date deposited: 01/07/2021
ISSN (electronic): 2666-8270
Publisher: Elsevier BV
Altmetrics provided by Altmetric