数据与计算发展前沿 ›› 2024, Vol. 6 ›› Issue (3): 67-72.

CSTR: 32002.14.jfdc.CN10-1649/TP.2024.03.007

doi: 10.11871/jfdc.issn.2096-742X.2024.03.007

• 会议论文 • 上一篇    下一篇

多波段多信使天文警报信息汇集与发布平台的设计与实现

王文帅1,*(),张红梅1,郑世界1,2,黄跃1,2,张正德1   

  1. 1.中国科学院高能物理研究所，北京 100049
    2.中国科学院粒子天体物理重点实验室，北京 100049
  • 收稿日期:2023-10-30 出版日期:2024-06-20 发布日期:2024-06-21
  • 通讯作者: *王文帅（E-mail: wangws@ihep.ac.cn）
  • 作者简介:王文帅，中国科学院高能物理研究所，高级工程师，博士，长期从事科研管理信息化平台建设，主要研究方向为数据管理、数据挖掘和信息化系统研发。
    负责论文初稿撰写与天文警报汇平台研发。
    WANG Wenshuai is a senior engineer at the Institute of High Energy Physics, Chinese Academy of Sciences.
    He has long been engaged in the construction of scientific research management information platforms. His research interests include data management, data mining, and information system development.
    In this paper, he is responsible for the paper drafting and system development.
    E-mail: wangws@ihep.ac.cn
  • 基金资助:
    中国科学院空间科学（二期）战略性先导科技专项(XDA15000000)

Design and Implementation of a Multi-Band and Multi-Messenger Astronomical Alert Information Collection and Publication Platform

WANG Wenshuai1,*(),ZHANG Hongmei1,ZHENG Shijie1,2,HUANG Yue1,2,ZHANG Zhengde1   

  1. 1. Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049, China
    2. Key Laboratory of Particle Astrophysics, Chinese Academy of Sciences, Beijing 100049, China
  • Received:2023-10-30 Online:2024-06-20 Published:2024-06-21

RichHTML

3

PDF

236

摘要：

【目的】近年来天文学研究进入多信使、多波段天文学时代，天文学家可以根据天文警报信息及时地调整观测计划，追踪观测热点事件，及时全面地获取天文警报信息非常重要。【方法】基于天文警报信息获取的需求，设计了天文警报汇平台，一站式汇集国际主流信息渠道数据，利用ChatGPT完成关键信息抽取，优化提示工程，提高ChatGPT信息抽取成功率，对警报数据融合整理后进行可视化展示。【结果】实验验证了提示工程的优化使ChatGPT抽取关键信息的成功率得到显著提升。【结论】天文警报汇平台可以给天文学家提供便利，及时地发布、共享天文警报信息。ChatGPT在很大程度上能够帮助科研人员从原始警报文本中抽取需要的关键信息，提示工程的优化使ChatGPT的输出更加可靠。

关键词: ChatGPT, GECAM, 天文警报, 信息提取

Abstract:

[Objective] In recent years, astronomy research has entered the era of multi-messenger and multi-band astronomy. It is very important for astronomers to adjust the observation plan in time according to the astronomical alert information, track the observation hot events, and obtain the astronomical alert information in a timely and comprehensive manner. [Methods] Based on the demand for astronomical alert information acquisition, an astronomical alert aggregation platform is designed to collect data from the main international information channels in one stop. ChatGPT is used to complete key information extraction, and then prompt engineering is adopted to improve the success rate of ChatGPT information extraction. The alert data is integrated and organized on the astronomical alert aggregation platform for visual display. [Results] The experiment verifies that the optimization of prompt engineering significantly improves the success rate of ChatGPT in extracting key information. [Conclusions] The astronomical alert aggregation platform can provide convenience for astronomers to publish and share astronomical alert information in a timely manner. ChatGPT can help researchers extract the key information needed from the original alarm text, and the optimization of the prompt engineering makes the output of ChatGPT more reliable.

Key words: ChatGPT, GECAM, astronomical alert, information extraction