BEAM: A First Benchmark for Microdata Entity Alignment with Knowledge Graphs

Key Points

• The research aims to establish a benchmark for aligning microdata with knowledge graphs to improve data interoperability.
• Developed the BEAM benchmark for evaluating entity alignment methods.
• Conducted experiments using various alignment algorithms on microdata.
• Evaluated performance based on predefined metrics for alignment accuracy.
• Demonstrated substantial improvement in alignment accuracy over existing methods.
• Achieved alignment accuracy rates averaging 85% across tested datasets.
• Identified specific algorithms that outperformed others in accuracy and efficiency.