This paper presents several approaches of converting food recipe data into graph data structures. We attempt to model these recipes using a directed rooted tree graph network. With this data structure we are able to give a standardized structure to food recipes - which in turn makes them tractable to a variety of machine learning algorithms. Previous attempts at analyzing food recipe data have viewed this problem fundamentally as a Natural Language Processing problem - this view however does not take advantage of the inherent structure already within food recipes. The goal of this paper is to begin the exploration of potential graph architectures which can be applied to food recipes - and hopefully extrapolated to processes in other domains which lend themselves easily modelable by these data structures.

Creator

• Carr, William

Contributors

• Li, Yanhua
• Kong, Xiangnan

Degree

• MS

Unit

• Data Science

Publisher

• Worcester Polytechnic Institute

Identifier

• etd-71276

Advisor

• Li, Yanhua

Defense date

• 2022-07-21

Year

• 2022

Date created

• 2022-08-02

Resource type

• Thesis

Source

• etd-71276

Rights statement

• In Copyright

Last modified

• 2022-12-09