Improving Interpretability via Explicit Word Interaction Graph Layer.
Arshdeep SekhonHanjie ChenAman ShrivastavaZhe WangYangfeng JiYanjun QiPublished in: CoRR (2023)
Keyphrases
- graph representation
- n gram
- multi layer
- graph theory
- random walk
- weighted graph
- graph theoretic
- human computer interaction
- directed acyclic graph
- graph structure
- word sense disambiguation
- co occurrence
- graph mining
- graph matching
- directed graph
- connected components
- keywords
- bipartite graph
- prediction accuracy
- text classification
- graph databases
- graph partitioning
- language model
- graph clustering
- protein interaction
- word recognition
- graph construction
- application layer
- middle layer