With budget outlooks for National Statistical Agencies (NSA) likely to be declining over time, NSAs need to look for smarter ways to deliver statistical services with reduced costs, whilst maintaining the same high level of quality of their statistical offerings. The advent of Big Data provides an opportunity for NSAs to have more competitive data acquisition strategies, and hence potentially more efficient services. However, Tam and Clarke [12] outline statistical issues, (e.g. business case, validity of statistical inference and data access and ownership etc.) that need to be addressed before a Big Data source can be used for the regular production of statistics; and Tam [11] gives the necessary conditions for statistical biases from Big Data to be ignored, and uses State Space Models for the analysis of satellite imagery to predict crop types and crop areas, as an alternative method to direct data collection.
展开▼
机译：随着国家统计局（NSA）的预算前景可能会随着时间的推移而下降，NSA需要寻找更明智的方式来以较低的成本提供统计服务，同时保持其统计产品的高质量水平。大数据的出现为国家安全局提供了一个更具竞争性的数据获取策略的机会，从而有可能提供更高效的服务。但是，Tam和Clarke [12]概述了统计问题（例如业务案例，统计推断的有效性以及数据访问和所有权等），然后才能使用大数据源进行常规统计。 Tam [11]提供了必要的条件，可以忽略大数据的统计偏差，并使用状态空间模型对卫星图像进行分析以预测作物类型和作物面积，作为直接收集数据的替代方法。
展开▼
