English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The flourishing of knowledge graph applications has driven the need forentity alignment (EA) across KGs. However, the heterogeneity of practical KGs,characterized by differing scales, structures, and limited overlappingentities, greatly surpasses that of existing EA datasets. This discrepancyhighlights an oversimplified heterogeneity in current EA datasets, whichobstructs a full understanding of the advancements achieved by recent EAmethods. In this paper, we study the performance of EA methods in practicalsettings, specifically focusing on the alignment of highly heterogeneous KGs(HHKGs). Firstly, we address the oversimplified heterogeneity settings ofcurrent datasets and propose two new HHKG datasets that closely mimic practicalEA scenarios. Then, based on these datasets, we conduct extensive experimentsto evaluate previous representative EA methods. Our findings reveal that, inaligning HHKGs, valuable structure information can hardly be exploited throughmessage-passing and aggregation mechanisms. This phenomenon leads to inferiorperformance of existing EA methods, especially those based on GNNs. Thesefindings shed light on the potential problems associated with the conventionalapplication of GNN-based methods as a panacea for all EA datasets.Consequently, in light of these observations and to elucidate what EAmethodology is genuinely beneficial in practical scenarios, we undertake anin-depth analysis by implementing a simple but effective approach: Simple-HHEA.This method adaptly integrates entity name, structure, and temporal informationto navigate the challenges posed by HHKGs. Our experiment results conclude thatthe key to the future EA model design in practice lies in their adaptabilityand efficiency to varying information quality conditions, as well as theircapability to capture patterns across HHKGs.

Toward Practical Entity Alignment Method Design: Insights from New Highly Heterogeneous Knowledge Graph Datasets