数字时代人口监测中的多源数据应用——现状及展望

Multi-source Data for Population Monitoring in the Digital Age:Current Situation and Prospects

人口监测是开展人口规划、统计、预测及预警等的基础性工作,是政府科学决策的基石。我国人口监测的传统数据仍存在漏报重报、质量不高、时效不强、属性不全、融合不深等问题,难以实施有效的人口精细化管理,也限制了监测数据在人口学研究中的广泛应用。随着信息技术的快速发展及移动终端的进一步普及,手机信令、互联网、遥感信息等一系列可直接或间接捕捉“人口行为”的新型数据大量涌现并应用于人口监测,与人口普查、调查等传统数据采集手段互为补充,有望通过交叉协同形成更加全面、更为实用的人口监测多源大数据基础。从人口监测的传统数据、新型数据及新老数据融合三方面入手,系统梳理了数字时代用于人口监测的多源数据。首先,通过梳理传统人口监测的分类及相关应用研究,总结了传统数据的主要优势、瓶颈与不足,进而明确了融合新型数据的需求靶点和必要性。其次,从数据分类及优缺点对比、应用研究等角度对新型数据的相关文献进行梳理,重点分析现阶段应用新型数据的机会和挑战。最后,通过总结传统数据与新型数据融合应用的优势,提炼了有待进一步研究的议题,为应用多源数据实现人口动态监测,推进流动人口统计信息化、规范化和精细化,支撑政府政策制定和人口学研究建立了理论基础。

Population monitoring is the basic work to support the party and the state to carry out multiple goals such as population planning,statistics,prediction and early warning,and is the cornerstone of scientific government decision-making.At present,the traditional data of population monitoring in China still have some problems,such as missing and rereporting,low quality,weak timeliness,incomplete attributes,and insufficient integration.It is difficult to implement effective and fine population management,and also limits the wide application of monitoring data in demographic research.With the rapid development of information technology and the further popularization of mobile terminals,a series of new data such as mobile phone signaling,Internet and remote sensing information that can directly or indirectly capture“population behavior”have emerged in large numbers and been applied to population monitoring,complementing traditional data collection methods such as census and survey.It is expected to form a more comprehensive and practical multi-source big data foundation for population monitoring through cross-collaboration.This paper systematically combs the multi-source data used for population monitoring in the digital era from three aspects:traditional data,new data and the fusion of traditional and new data.Firstly,by combing the classification of traditional population monitoring and related application research,it summarizes the main advantages,bottlenecks and shortcomings of traditional data,and then clarifies the demand targets and necessity of fusion of new data.Secondly,from the perspective of data classification,comparison of advantages and disadvantages,and application research,the relevant literature of new data is reviewed,focusing on the opportunity and challenge of applying new data at the present stage.Finally,by summarizing the advantages of fusion application of traditional data and new data,the issues to be further studied are refined.This study establishes a theoretical basis for using multi-source data to realize population dynamic monitoring,promote the informationization,standardization and refinement of floating population statistics,and support government policy making and demographic research.