Knowledge Graph embedding based missing metadata prediction

The increasing volume and complexity of data collected in experimental settings necessitate rigorous methodologies to ensure that such metadata remains Findable, Accessible, Interoperable, and Reusable (FAIR). The RDF graph, as a representation of knowledge graph, can support these FAIR principles. However, metadata records accompanying legacy data may be incomplete for various reasons, such as adherence to outdated standards, omission of essential parameters and errors. To improve the quality of metadata, we employ knowledge graph embedding techniques to predict missing metadata stored in RDF graph. Subsequently, using a classifier to verify the accuracy of the predicted metadata. The best results show that our methods can correctly recover 89.5% of the metadata, with a precision of 93% after classification.