.png){kind=icon}
.png){kind=icon}
Return
Aspect-based sentiment analysis based on multi-granularity graph convolutional network
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neunet.2025.107864.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• We proposed a novel multi-granularity graph convolutional network for ABSA. • We designed a hierarchical constituency tree mask matrices. • We eliminated unnecessary layers in the constituency tree. • We fused the dependency and constituency trees based on structural features.
Keywords:
multi-granularity graph convolutional network
hierarchical constituency tree mask matrices
constituency tree
dependency tree
structural features
Journal
IF:
6.3
Papers:
7.8K
Citations:
3.0W