Leveraging digital news to create databases of the impacts of small, medium and, large disasters

This project consisted in the development of protocols to systematically collect, organize, analyze and tabulate information related to the impact of disasters of all scales (small, medium, and large) from printed and digital newspapers, particularly in places where the study of the impacts of disasters is neglected despite their high frequency of occurrence (e.g., Guatemala).
Disaster impact measures were characterized in terms of human lives (i.e., deaths, injured, missing, directly and indirectly, affected, evacuated, relocated), economic (i.e., USD dollars and local currency), damage to infrastructure (e.g., education centers, hospitals, roads, bridges, energy, water supply, and sewage systems), industry (in Central America, mainly agricultural and livestock), different types of disruption caused by the disaster (e.g., school closure days, community isolation time), aid, and many others. The database includes measures of disaster impacts in Guatemala for the period from 2015 to 2021.
The methodology for creating disasters dataset from printed and digital news media is not new and has been used by many multidisciplinary researchers worldwide. However, this project is unique in that it focuses mainly on streamlining archival processes through digital media. The data, instruments, and methodology developed in this project can help multiple researchers facilitate the processes related to digital archival methods as they relate to the occurrence of disasters and their impacts, particularly on locations where the study of small, medium, and large disasters is neglected. Thus, multidisciplinary natural hazard and disaster researchers will be the immediate beneficiaries of the dataset.
Generating data from digital media has the potential to improve the quality, quantity, and speed of the study of disasters, in addition to other benefits such as an increase in victim-centered approaches and public awareness, or the possibility of including impacts on historically marginalized populations that have more frequent interactions with disasters of different scales and type.
The final products of this project are specifically three different protocols to collect newspaper information from automated and manual methods, and one database of disaster impacts developed from analyzing digital news media publications.

本项目旨在开发系统化的采集、整理、分析与制表流程，用于从印刷及数字报纸中收集各类规模（小型、中型、大型）灾害的影响相关信息，尤其聚焦于那些尽管灾害发生频率极高，但相关灾害影响研究仍被忽视的地区（例如危地马拉）。
灾害影响评估指标涵盖多维度：人员维度包括遇难、受伤、失踪、直接与间接受影响人口、疏散及安置人口；经济维度涵盖以美元（USD）及当地货币计价的损失；基础设施破坏涵盖教育机构、医院、道路、桥梁、能源、供水及排污系统等；产业维度（中美洲地区主要为农牧业）；灾害引发的各类中断情形，例如停课天数、社区隔离时长；以及援助相关数据等诸多方面。本数据库收录了2015至2021年间危地马拉境内的灾害影响评估数据。
从印刷与数字新闻媒体构建灾害数据集的方法论并非首创，全球范围内已有众多多学科研究者采用过类似方法。但本项目的独特之处在于，其核心聚焦于通过数字媒体优化归档流程。本项目开发的数据集、工具与方法论，可助力全球多领域研究者简化与灾害发生及其影响相关的数字归档工作流程，尤其针对那些中小型及大型灾害研究被忽视的地区。因此，多学科自然灾害与灾害研究人员将成为本数据集的首批受益者。
从数字媒体采集数据，不仅能够提升灾害研究的质量、覆盖范围与研究速度，还具备其他诸多优势，例如推动以受灾群体为中心的研究范式、提升公众防灾意识，或是能够纳入对历史上处于边缘地位、更频繁遭遇各类规模与类型灾害的群体的影响评估。
本项目的最终成果具体包括三套分别采用自动化与人工方式采集报纸信息的操作规范，以及一套通过分析数字新闻媒体报道构建的灾害影响数据库。