Using Word Embeddings to Learn a Better Food Ontology

Youn, Jason; Naravane, Tarini; Tagkopoulos, Ilias

doi:10.3389/frai.2020.584784

Cited by 12 publications

(4 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to expand the food word dictionary by discovering new words, we follow the assumption that food words with similar characteristics will exhibit certain patterns in the word embedding space [ 79 ]. This means that similar words can be clustered together and form clusters in the word embedding space based on the different food groups they belong to.…”

Section: Methodsmentioning

confidence: 99%

Empowering health geography research with location-based social media data: innovative food word expansion and energy density prediction via word embedding and machine learning

Wang,

Kim,

Chang

2023

Int J Health Geogr

View full text Add to dashboard Cite

Background The exponential growth of location-based social media (LBSM) data has ushered in novel prospects for investigating the urban food environment in health geography research. However, previous studies have primarily relied on word dictionaries with a limited number of food words and employed common-sense categorizations to determine the healthiness of those words. To enhance the analysis of the urban food environment using LBSM data, it is crucial to develop a more comprehensive list of food-related words. Within the context, this study delves into the exploration of expanding food-related words along with their associated energy densities. Methods This study addresses the aforementioned research gap by introducing a novel methodology for expanding the food-related word dictionary and predicting energy densities. Seed words are generated from official and crowdsourced food composition databases, and new food words are discovered by clustering food words within the word embedding space using the Gaussian mixture model. Machine learning models are employed to predict the energy density classifications of these food words based on their feature vectors. To ensure a thorough exploration of the prediction problem, ten widely used machine learning models are evaluated. Results The approach successfully expands the food-related word dictionary and accurately predicts food energy density (reaching 91.62%.). Through a comparison of the newly expanded dictionary with the initial seed words and an analysis of Yelp reviews in the city of Toronto, we observe significant improvements in identifying food words and gaining a deeper understanding of the food environment. Conclusions This study proposes a novel method to expand food-related vocabulary and predict the food energy density based on machine learning and word embedding. This method makes a valuable contribution to building a more comprehensive list of food words that can be used in geography and public health studies by mining geotagged social media data.

show abstract

Section: Methodsmentioning

confidence: 99%

Empowering health geography research with location-based social media data: innovative food word expansion and energy density prediction via word embedding and machine learning

Wang,

Kim,

Chang

2023

Int J Health Geogr

View full text Add to dashboard Cite

show abstract

“…The team has developed key resources such as the SNAPMe Benchmark Database for the evaluation of algorithms for computer vision in dietary assessment and the Food Atlas, a KnowledgeBase that uses deep learning, natural language processing, large language models, and other techniques to connect foods, ingredients, compounds, and health effects. Applications include the prediction of food composition after processing (Naravane & Tagkopoulos, 2023), algorithmic food matching and tolerance methods (Eetemadi & Tagkopoulos, 2023), automated ontology construction for food (Youn, Naravane, & Tagkopoulos, 2020), and causal prediction of dietary intervention efficacy in digestive diseases (Eetemadi & Tagkopoulos, 2021).…”

Section: Research Clustersmentioning

confidence: 99%

The AIFS Institute: Building a better food system through AI

Tagkopoulos,

Earles,

Lemay

et al. 2024

AI Magazine

Self Cite

View full text Add to dashboard Cite

Our food system is complex, multifaceted, and in need of an upgrade. Population growth, climate change, and socioeconomic disparities are some of the challenges that create a systemic threat to its sustainability and capacity to address the needs of an evolving planet. The mission of the AI Institute of Next Generation Food Systems (AIFS) is to leverage the latest advances in AI to help create a more sustainable, efficient, nutritious, safe, and resilient food system. Instead of using AI in isolation, AIFS views it as the connective tissue that can bring together interconnected solutions from farm to fork. From guiding molecular breeding and building autonomous robots for precision agriculture, to predicting pathogen outbreaks and recommending personalized diets, AIFS projects aspire to pave the way for infrastructure and systems that empower practitioners to build the food system of the next generation. Workforce education, outreach, and ethical considerations related to the emergence of AI solutions in this sector are an integral part of AIFS with several collaborative activities aiming to foster an open dialogue and bringing closer students, trainees, teachers, producers, farmers, workers, policy makers, and other professionals.

show abstract

“…Key steps necessary for developing a food ontology include the following: (1) identifying the scope and purpose of an ontology (i.e., is it for a specific health condition such as prediabetes, or it is for general wellness); (2) identifying and importing appropriate classes from existing reference ontol-ogies (e.g., FOBI, 84 FoodOn, 80 and others 85,86 ); and (3) creating new classes/relationships for any conceptualization required for models and themes within the previously identified scope and purpose but not found in existing ontologies. Ontologies to facilitate the workflow shown in ►Fig.…”

Section: Representing Food Practicementioning

confidence: 99%

Opportunities and Challenges of Integrating Food Practice into Clinical Decision-Making

2022

View full text Add to dashboard Cite

Background Food practice plays an important role in health. Food practice data collected in daily living settings can inform clinical decisions. However, integrating such data into clinical decision-making is burdensome for both clinicians and patients, resulting in poor adherence and limited utilization. Automation offers benefits in this regard, minimizing this burden resulting in a better fit with a patient's daily living routines, and creating opportunities for better integration into clinical workflow. Although the literature on patient-generated health data (PGHD) can serve as a starting point for the automation of food practice data, more diverse characteristics of food practice data provide additional challenges. Objectives We describe a series of steps for integrating food practices into clinical decision-making. These steps include the following: (1) sensing food practice; (2) capturing food practice data; (3) representing food practice; (4) reflecting the information to the patient; (5) incorporating data into the EHR; (6) presenting contextualized food practice information to clinicians; and (7) integrating food practice into clinical decision-making. Methods We elaborate on automation opportunities and challenges in each step, providing a summary visualization of the flow of food practice-related data from daily living settings to clinical settings. Results We propose four implications of automating food practice hereinafter. First, there are multiple ways of automating workflow related to food practice. Second, steps may occur in daily living and others in clinical settings. Food practice data and the necessary contextual information should be integrated into clinical decision-making to enable action. Third, as accuracy becomes important for food practice data, macrolevel data may have advantages over microlevel data in some situations. Fourth, relevant systems should be designed to eliminate disparities in leveraging food practice data. Conclusion Our work confirms previously developed recommendations in the context of PGHD work and provides additional specificity on how these recommendations apply to food practice.

show abstract

Using Word Embeddings to Learn a Better Food Ontology

Cited by 12 publications

References 30 publications

Empowering health geography research with location-based social media data: innovative food word expansion and energy density prediction via word embedding and machine learning

Empowering health geography research with location-based social media data: innovative food word expansion and energy density prediction via word embedding and machine learning

The AIFS Institute: Building a better food system through AI

Opportunities and Challenges of Integrating Food Practice into Clinical Decision-Making

Contact Info

Product

Resources

About