FC-KBQA: A Fine-to-Coarse Composition Framework for Knowledge Base Question Answering

TL;DR

FC-KBQA introduces a fine-to-coarse framework for KBQA that enhances generalization and executability by reformulating fine-grained KB components into middle-grained pairs, achieving state-of-the-art results and faster performance.

Contribution

The paper presents a novel FC-KBQA framework that improves KBQA by combining fine- and coarse-grained knowledge modeling for better generalization and efficiency.

Findings

Achieves state-of-the-art performance on GrailQA and WebQSP datasets.

Runs four times faster than baseline methods.

Effectively balances generalization and executability in KBQA.

Abstract

The generalization problem on KBQA has drawn considerable attention. Existing research suffers from the generalization issue brought by the entanglement in the coarse-grained modeling of the logical expression, or inexecutability issues due to the fine-grained modeling of disconnected classes and relations in real KBs. We propose a Fine-to-Coarse Composition framework for KBQA (FC-KBQA) to both ensure the generalization ability and executability of the logical expression. The main idea of FC-KBQA is to extract relevant fine-grained knowledge components from KB and reformulate them into middle-grained knowledge pairs for generating the final logical expressions. FC-KBQA derives new state-of-the-art performance on GrailQA and WebQSP, and runs 4 times faster than the baseline.