Capturing Food Knowledge with Semantics and Embeddings

Anonymous

Nov 17, 2018 AKBC 2019 Conference Blind Submission readers: everyone Show Bibtex
  • Keywords: data acquisiton, semantic extract transform load, concept similarity, entity resolution, relationship discovery, nutrition
  • Abstract: Poor quality eating patterns contribute significantly to the incidence of preventable chronic diseases. The proliferation of recipes and other food information sources on the Web presents an opportunity for discovering and organizing diet related knowledge into a knowledge graph, which can in turn be used to generate food recommendations tailored to an individual's dietary habits and preferences. In this paper, we present our work on building a food knowledge graph using semantics oriented knowledge ingestion. We further augment the knowledge graph by incorporating inferences we have made on the similarity between food ingredients. These similarities are derived from embeddings generated from online recipe data. In the true spirit of linked data, we have linked to many of the existing concepts in other related ontologies, as well as community maintained resources such as DBpedia. The resulting knowledge graph is capable of answering questions related to the composition of dishes, nutritional content of food items and potential substitutions.
  • Archival status: Archival
  • Subject areas: Natural Language Processing, Information Extraction, Information Integration, Knowledge Representation, Semantic Web
  • TL;DR: Knowledge graph for food recommendation
0 Replies

Loading