Improving Complex Knowledge Base Question Answering via Question-to-Action and Question-to-Question Alignment

TL;DR

This paper introduces ALCQA, a framework that improves complex knowledge base question answering by aligning questions with actions and similar questions, leading to significant performance gains.

Contribution

The paper proposes a novel alignment-enhanced framework for complex KBQA, incorporating question-to-action and question-to-question alignment to bridge semantic gaps.

Findings

Outperforms state-of-the-art methods on CQA and WQSP datasets.

Achieves a 9.88% improvement in F1 score on CQA dataset.

Demonstrates effectiveness of alignment strategies in KBQA.

Abstract

Complex knowledge base question answering can be achieved by converting questions into sequences of predefined actions. However, there is a significant semantic and structural gap between natural language and action sequences, which makes this conversion difficult. In this paper, we introduce an alignment-enhanced complex question answering framework, called ALCQA, which mitigates this gap through question-to-action alignment and question-to-question alignment. We train a question rewriting model to align the question and each action, and utilize a pretrained language model to implicitly align the question and KG artifacts. Moreover, considering that similar questions correspond to similar action sequences, we retrieve top-k similar question-answer pairs at the inference stage through question-to-question alignment and propose a novel reward-guided action sequence selection strategy to select from candidate action sequences. We conduct experiments on CQA and WQSP datasets, and the results show that our approach outperforms state-of-the-art methods and obtains a 9.88\% improvements in the F1 metric on CQA dataset. Our source code is available at https://github.com/TTTTTTTTy/ALCQA.