Skip navigation
Please use this identifier to cite or link to this item:
Title: Sous Chef: An Automated Recipe Ingredient Tagger
Authors: Yuen, Michelle
Advisors: Levy, Amit
Department: Computer Science
Class Year: 2020
Abstract: This paper documents a novel approach to ingredient tagging through the use of regular expressions to separate parts of ingredient text into quantity, unit, comment, and name. Previous approaches to automated ingredient tagging relied on natural language processing or building complex machine learning models. Instead, Sous Chef uses regular expressions to parse ingredient text. However, the problem of ingredient tagging comes with the added difficulty of ambiguity when it comes to tagging content. Additionally, many automated taggers struggle with the fact that ingredient text can vary wildly in structure, and cannot be entirely assumed, making it difficult to achieve high accuracy across a variety of data. To address these two issues of ambiguity and lack of structure, this paper focuses on the creation of a web application where users can parse ingredients, manually tag ingredients, and fix ingredient text to better cooperate with automated taggers. This paper details the process behind creating a new automated tagger, and the implementation of a web application to aid further evaluation of the automated ingredient tagging problem.
Type of Material: Princeton University Senior Theses
Language: en
Appears in Collections:Computer Science, 1988-2020

Files in This Item:
File Description SizeFormat 
YUEN-MICHELLE-THESIS.pdf888.33 kBAdobe PDF    Request a copy

Items in Dataspace are protected by copyright, with all rights reserved, unless otherwise indicated.