Conversational Question Answering with Reformulations over Knowledge Graph

TL;DR

This paper introduces CornNet, a reinforcement learning-based model that uses large language models to generate reformulations of questions, significantly improving conversational question answering over knowledge graphs.

Contribution

The paper presents a novel RL-based approach with a teacher-student architecture leveraging LLM-generated reformulations to enhance convQA performance.

Findings

CornNet outperforms existing convQA models in experiments.

Reformulations generated by LLMs improve question interpretation.

The teacher-student architecture effectively transfers reformulation knowledge.

Abstract

Conversational question answering (convQA) over knowledge graphs (KGs) involves answering multi-turn natural language questions about information contained in a KG. State-of-the-art methods of ConvQA often struggle with inexplicit question-answer pairs. These inputs are easy for human beings to understand given a conversation history, but hard for a machine to interpret, which can degrade ConvQA performance. To address this problem, we propose a reinforcement learning (RL) based model, CornNet, which utilizes question reformulations generated by large language models (LLMs) to improve ConvQA performance. CornNet adopts a teacher-student architecture where a teacher model learns question representations using human writing reformulations, and a student model to mimic the teacher model's output via reformulations generated by LLMs. The learned question representation is then used by an RL model to locate the correct answer in a KG. Extensive experimental results show that CornNet outperforms state-of-the-art convQA models.