开发贫困和脆弱性评估的数据集成工具：菲律宾试点倡议的反思（英）

Developing a Data IntegrationTool for Poverty and Vulnerability KEY POINTS •In many developingeconomies, there is aneed to improve inclusiveurban planning, disasterrisk reduction, and Mariko ShibasakiStrategic Initiative ConsultantLocationMind Inc. Mar Andriel UmaliConsultant •Combining GlobalPositioning System data,census data, satellite imagery,and survey responses intoa user-friendly platform,a data integration toolcalled Poverty Impact andVulnerability Evaluation orPIVE was pilot-testedin select areas in the Arturo Martinez, Jr.Senior StatisticianADB Hiroyuki MiyazakiChief Executive Officer,GLODAL, Inc. Deepanshu AgarwalStrategic Initiative ConsultantLocationMind Inc. Ryosuke ShibasakiChief Technology Officer,LocationMind Inc., and Saurav RanjitSenior Data EngineerLocationMind Inc. Hong Soo LeeSenior Urban Development Specialist The PIVE tool is alsointended to strengthendisaster risk reductionefforts by integrating hazardmaps with settlement data INTRODUCTION The exponential increase in data creation is one of the defining features of ourcontemporary era, with estimates suggesting that more than 400 million terabytes of Results from stakeholderconsultations suggest thatPIVE provides detailed datainsights and a user-friendlyinterface, highlighting thetool’s potential for broader Notes: In this publication, “$” refers to United States dollars and “₱” refers to Philippine pesos. This briefwas prepared under Asian Development Bank (ADB) Technical Assistance (TA) 6926: Key Indicatorsfor Asia and the Pacific 2021–2023 Subproject 3 and ADB TA 6856: Development of New StatisticalResources and Building Capacity in New Data Sources and Technologies. The authors would like to expresstheir gratitude to all who contributed to this work. Sincere appreciation goes to the dedicated ADB staffin the PIVE initiative for their invaluable assistance: Josephine Aquino (Department of Communicationsand Knowledge Management), Emeli N. Noller (consultant for Department of Communications and ISBN 978-92-9277-574-2 (print)ISBN 978-92-9277-575-9 (PDF)ISSN 2071-7202 (print)ISSN 2218-2675 (PDF)Publication Stock No. BRF250516-2 ADB BRIEFS NO. 372 For instance, monitoring progress toward the SustainableDevelopment Goals (SDGs) necessitates highly granular datadisaggregated by income, sex, age, race, ethnicity, migration status,disability, geographic location, and other relevant characteristicsto ensure that no one is left behind. Traditionally, data used formonitoring progress with respect to various development targetshave been gathered through sample surveys, censuses, andadministrative sources. However, the advent of digital data from designed to assist policymakers, planners, and other stakeholdersin disaster management and urban planning. The innovative design of PIVE automates and visualizes dataintegration on a dashboard for development practitioners andpolicymakers who may not have the technical expertise inmerging heterogeneous datasets. It specifically illustrates howto identify users’ data-analytic goals and addresses semanticheterogeneity to identify relevant datasets. By enabling the The integrated approach adopted by PIVE allows for thedisaggregation of data by critical variables such as socioeconomicstatus, age, gender, ethnicity, and geographic location.Consequently, it can reveal nuanced disparities in the impacts ofemergencies and disasters that might otherwise remain obscuredwhen examining individual datasets separately. By highlightinghow marginalized communities may be disproportionately The coronavirus disease (COVID-19) pandemic underscoredthe critical need for integrating multiple datasets as combiningofficial statistics with Big Data can potentially provide moretimely and effective responses particularly in a time of acrisis. During the pandemic, official statistics compiled fromtraditional data sources such as health records, census data,and administrative reports were essential, but often insufficientfor real-time decision-making. The rapid spread of the virusrequired immediate and precise information to track infectionrates, allocate resources, and implement public health measures.By integrating Big Data sources—mobile phone records, socialmedia activity, and satellite imagery—with official statistics,authorities were able to gain a more comprehensive and dynamicunderstanding of the pandemic’s progression. For instance,examination of mobile phone data helped track human mobilitypatterns, which were crucial for understanding how the virusspread in different regions (McGough et al. 2020). In general, Through a structured evaluation in the context of pilot-testingthe PIVE tool in select areas in the Philippines, the PIVE tooldemonstrates strong performance in capturing varied and dynamicdata-analytic goals and generating integrated data tailored tothese goals. This initiative provides both operational lessons Data IntegrationData integration is an emerging statistical frontier that nat