Abstract:Shallow semantic parsing assigns a simple structure (such as WHO did WHAT to WHOM, WHEN, WHERE, WHY, and HOW) to each predicate in a sentence. It plays a critical role in event-based information extraction and thus is important for deep information processing and management. This paper proposes a tree kernel method for a particular shallow semantic parsing task, called semantic role labeling (SRL), with an enriched parse tree structure. First, a new tree kernel is presented to effectively capture the inherent structured knowledge in a parse tree by enabling the standard convolution tree kernel with context-sensitiveness via considering ancestral information of substructures and approximate matching via allowing insertion/deletion/substitution of tree nodes in the substructures. Second, an enriched parse tree structure is proposed to both well preserve the necessary structured information and effectively avoid noise by differentiating various portions of the parse tree structure. Evaluation on the CoNLL'2005 shared task shows that both the new tree kernel and the enriched parse tree structure contribute much in SRL and our tree kernel method significantly outperforms the state-of-the-art tree kernel methods. Moreover, our tree kernel method is proven rather complementary to the state-of-the-art feature-based methods in that it can better capture structural parse tree information.

Forest-Based Semantic Role Labeling

Semantic Role Labeling as Dependency Parsing: Exploring Latent Tree Structures inside Arguments.

An MRC Framework for Semantic Role Labeling

Tree kernel-based semantic role labeling with enriched parse tree structure

Fast and Accurate Span-based Semantic Role Labeling As Graph Parsing

Syntax-aware Neural Semantic Role Labeling

Semantic Role Labeling as Syntactic Dependency Parsing

Research on Chinese Semantic Role Labeling Based on Shallow Parsing

Chinese Semantic Role Labeling with Shallow Parsing.

Semantic Role Labeling Based on Dependency Relationship

An integrated strategy for Chinese semantic role labeling

Research on Chinese Semantic Role Labeling Based on Ocean Big Data

Fast and Accurate End-to-End Span-based Semantic Role Labeling as Word-based Graph Parsing.

Semantic Role Labeling for Learner Chinese: the Importance of Syntactic Parsing and L2-L1 Parallel Data

Semantic Role Labeling with Heterogeneous Syntactic Knowledge

Deep semantic role labeling: What works and what’s next

Semantic Role Labeling Based on Dependency Tree with Multi-features

Semantics-Driven Shallow Parsing for Chinese Semantic Role Labeling.

Semantic Role Labeling of Clinical Text: Comparing Syntactic Parsers and Features.

Syntax-Enhanced Self-Attention-Based Semantic Role Labeling

A Full End-to-End Semantic Role Labeler, Syntax-agnostic over Syntax-aware?