Food is an integral part of everyday life, and food choices directly affect one’s health. Both academics and practitioners have attempted to help consumers make good decisions about their food choices and recommended better or healthier alternatives. However, in thinking about food it is important to put it in context, as each food item is often combined with other food items to create the gestalt of a recipe or meal. Understanding the complex interaction between food items that are used or consumed together is crucial to provide effective recommendations.
In this research, I leverage tools from machine learning and textual analysis like the embedding approach for representation learning to understand food in its context and to build recommender systems that account for the complementarity or fit of co-consumed food items. I show that this consideration of fit among food items can lead to better and healthier food recommendations.
Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/ycr7-n633 |
Date | January 2023 |
Creators | Sozuer Zorlu, Sibel |
Source Sets | Columbia University |
Language | English |
Detected Language | English |
Type | Theses |
Page generated in 0.0018 seconds