Conversational Question Answering with Reformulations over Knowledge Graph

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Lihui, Hill, Blaine, Du, Boxin, Wang, Fei, Tong, Hanghang
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914732764561408
author Liu, Lihui
Hill, Blaine
Du, Boxin
Wang, Fei
Tong, Hanghang
author_facet Liu, Lihui
Hill, Blaine
Du, Boxin
Wang, Fei
Tong, Hanghang
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2312_17269
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Conversational Question Answering with Reformulations over Knowledge Graph
Liu, Lihui
Hill, Blaine
Du, Boxin
Wang, Fei
Tong, Hanghang
Computation and Language
Artificial Intelligence
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.
title Conversational Question Answering with Reformulations over Knowledge Graph
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2312.17269