What is it about?
QUERY2BERT is a model that is used to reasoning the answer for logical queries over knowledge graph. It solves missing link over KG using node2vec and BERT to embed whole KG. Then, embedded nodes were indexed by K-D tree to retrieve fast. Finally, it defines logical operators like projection, union, intersection and negation.
Featured Image
Photo by julien Tromeur on Unsplash
Why is it important?
QUERY2BERT solves missing link over knowledge graph to answer logical queries that require multi-hop reasoning.
Perspectives
Writing this article was a great pleasure as it has co-authors with whom I have had long standing collaborations. This article also bring the reasoning of human into machine and ultimately to a greater involvement in artificial intelligence reasoning research.
Phan H. V. Trường
Read the Original
This page is a summary of: QUERY2BERT: Combining Knowledge Graph and Language Model for Reasoning on Logical Queries, IEEE Access, January 2025, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/access.2025.3528097.
You can read the full text:
Contributors
The following have contributed to this page
