New method fuses graph and text for patent entity alignment

This paper introduces a novel method for aligning entities within science and technology patent knowledge graphs. The proposed approach leverages a graph convolution network combined with the BERT model to fuse structural information from the graph with textual attributes like patent names and descriptions. This multi-information fusion aims to improve the accuracy of entity alignment, outperforming existing methods on benchmark datasets according to Hits@K evaluation metrics. AI

IMPACT This research could improve the organization and retrieval of patent information, potentially aiding innovation and intellectual property management.

RANK_REASON The cluster contains an academic paper detailing a new method for entity alignment in patent data. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• entity alignment
• Hits@K
• knowledge graph
• science and technology patent
• Zeli Guan

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →