Multi-Modal Entity Alignment Method Based on Feature Enhancement

Multi-modal entity alignment refers to identifying equivalent entities between two different multi-modal knowledge graphs that consist of information such as structural triples and descriptive images. Most previous methods have mainly used corresponding encoders each modality encode then perform feature fusion obtain the joint representation. However, this approach does not fully utilize aligned entities. To address issue, we propose MEAFE, a method based on enhancement. The MEAFE adopts pre-trained model, OCR GATv2 network enhance model’s ability extract useful features in structure triplet image description, respectively, thereby generating more effective representations. Secondly, it further adds modal distribution understanding modeling information. Experiments bilingual cross-graph datasets demonstrate proposed outperforms models use traditional extraction methods.