Dual Hypergraph Features for Path Inference in Wikipedia Links | IEEE Conference Publication | IEEE Xplore

Dual Hypergraph Features for Path Inference in Wikipedia Links


Abstract:

Here, path extrapolation is assessed in a graph by extracting the most informative features from nodes and edges. Wikipedia articles are the graph nodes, and the links be...Show More

Abstract:

Here, path extrapolation is assessed in a graph by extracting the most informative features from nodes and edges. Wikipedia articles are the graph nodes, and the links between articles are the graph edges. A graph neural network called GRETEL is used as baseline. New features are extracted by exploiting the information from the data. By employing the dual hypergraph transformation the structural role of nodes and edges is interchanged, enabling more complex relationships to be captured. Experimental evidence is disclosed demonstrating that the features extracted from the dual hypergraph are more de-scriptive than those extracted from the original graph, improving GRETEL performance.
Date of Conference: 18-23 June 2023
Date Added to IEEE Xplore: 02 August 2023
ISBN Information:

ISSN Information:

Conference Location: Gold Coast, Australia

Funding Agency:


References

References is not available for this document.