Leveraging Semantic Parsing for Relation Linking over Knowledge Bases

TL;DR

This paper introduces SLING, a relation linking framework that uses semantic parsing and distant supervision to improve relation extraction for knowledgebase question answering, addressing language ambiguity and training data scarcity.

Contribution

SLING combines multiple relation linking strategies leveraging semantic parsing and distant supervision, achieving state-of-the-art results on multiple KBQA datasets.

Findings

Achieves state-of-the-art performance on QALD-7, QALD-9, and LC-QuAD 1.0 datasets.

Effectively addresses language ambiguity and training data limitations.

Integrates linguistic cues, semantic representations, and knowledgebase information.

Abstract

Knowledgebase question answering systems are heavily dependent on relation extraction and linking modules. However, the task of extracting and linking relations from text to knowledgebases faces two primary challenges; the ambiguity of natural language and lack of training data. To overcome these challenges, we present SLING, a relation linking framework which leverages semantic parsing using Abstract Meaning Representation (AMR) and distant supervision. SLING integrates multiple relation linking approaches that capture complementary signals such as linguistic cues, rich semantic representation, and information from the knowledgebase. The experiments on relation linking using three KBQA datasets; QALD-7, QALD-9, and LC-QuAD 1.0 demonstrate that the proposed approach achieves state-of-the-art performance on all benchmarks.