arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2404.01720cs.CL
收藏

面向时序知识图谱问答的自改进编程方法

Self-Improvement Programming for Temporal Knowledge Graph Question Answering

Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所);University of Chinese Academy of Sciences(中国科学院大学)

Zhuo Chen, Zhao Zhang, Zixuan Li, Fei Wang, Yutao Zeng, Xiaolong Jin, Yongjun Xu

AI总结:

本文提出 Prog-TQA,通过设计时序运算符并利用大语言模型上下文学习生成可执行程序,结合自改进策略提升时序知识图谱问答性能,在 MultiTQ 和 CronQuestions 上取得领先结果。

摘要:

时序知识图谱问答(TKGQA)旨在回答针对时序知识图谱(TKGs)具有时间意图的问题。该任务的核心挑战在于理解问题中涉及多种时间约束类型(例如 before、first)的复杂语义信息。现有的端到端方法通过学习问题和候选答案的时间感知嵌入来隐式建模时间约束,远未达到对问题的全面理解。受基于语义解析的方法启发,这类方法通过生成带有符号运算符的逻辑形式来显式建模问题中的约束,我们为时间约束设计了基础时序运算符,并提出了一种新颖的面向 TKGQA 的自改进编程方法(Prog-TQA)。具体而言,Prog-TQA 利用大语言模型(LLMs)的上下文学习能力来理解问题中的组合时间约束,并在给定少量示例的情况下生成相应的程序草稿。然后,它通过链接模块将这些草稿与 TKGs 对齐,并随后执行它们以生成答案。为了增强理解问题的能力,Prog-TQA 进一步配备了自改进策略,利用高质量的自生成草稿有效地引导 LLMs。大量实验证明了所提出的 Prog-TQA 在 MultiTQ 和 CronQuestions 数据集上的优越性,尤其是在 Hits@1 指标上。

英文摘要:

Temporal Knowledge Graph Question Answering (TKGQA) aims to answer questions with temporal intent over Temporal Knowledge Graphs (TKGs). The core challenge of this task lies in understanding the complex semantic information regarding multiple types of time constraints (e.g., before, first) in questions. Existing end-to-end methods implicitly model the time constraints by learning time-aware embeddings of questions and candidate answers, which is far from understanding the question comprehensively. Motivated by semantic-parsing-based approaches that explicitly model constraints in questions by generating logical forms with symbolic operators, we design fundamental temporal operators for time constraints and introduce a novel self-improvement Programming method for TKGQA (Prog-TQA). Specifically, Prog-TQA leverages the in-context learning ability of Large Language Models (LLMs) to understand the combinatory time constraints in the questions and generate corresponding program drafts with a few examples given. Then, it aligns these drafts to TKGs with the linking module and subsequently executes them to generate the answers. To enhance the ability to understand questions, Prog-TQA is further equipped with a self-improvement strategy to effectively bootstrap LLMs using high-quality self-generated drafts. Extensive experiments demonstrate the superiority of the proposed Prog-TQA on MultiTQ and CronQuestions datasets, especially in the Hits@1 metric.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑