Relationships Discovery between Nutritional Disorders and Gut Microbiota with Knowledge Graphs

Abstract

Nutrition is closely associated with public health, and more and more studies have demonstrated that there are many correlations between the gut microbiota and nutritional disorders. These studies constitute the basic source of medical knowledge. However, discovering knowledge from massive amounts of data is not easy only by manpower. This study aims at discovering the relationships between nutritional disorders and gut microbiota from the biomedical literatures with knowledge graph. In this study, literature content information from massive data is automatically and instantly extracted using SPARQL query and logic reasoning technologies. Then, an annotated corpus is obtained with the categorization and definition of relationships between nutritional disorders and gut microbiota. Thus, a knowledge graph of semantic relationship between nutritional disorders and gut microbiota is constructed, and semantic relationships are discovered, verifying the feasibility and high efficiency of relationships discovery with knowledge graphs. Moreover, the analysis of evidence case shows that it is necessary and valuable to use the importance of evidence to exclude conflicting conclusions.