Scholarlink

Over the past decade, National Italian Institute of Statistics (Istat) has recognized the potential of big data and machine learning (ML) in modernizing statistical production. This chapter outlines Istat’s journey from early experimentation to new statistical production pipelines based on Trusted Smart Statistics (TSS) principles. We discuss integrating big data sources—such as satellite imagery, web data, and sensor-based information—into official statistics and the challenges this paradigm shift poses. The chapter describes several research projects, including maritime traffic analysis using automatic identification system (AIS) data, urban green area estimation through remote sensing, and web intelligence for business statistics. Special attention is given to the methodological issues surrounding ML applications in official statistics, particularly in relation to estimation, imputation, and quality assessment. We examine the evolving role of data science skills within Istat, the institutional investments in human capital, and involvement in international research projects that have supported innovation. By integrating ML and big data into statistical processes, Istat aims to enhance data quality, improve timeliness, and provide more granular insights while ensuring compliance with statistical principles and data governance frameworks. The chapter concludes with considerations on future challenges and opportunities for official statistics in the datafied society.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Big Data and Machine Learning at Istat

  • 章节

  • 出版模式：

    开放获取

  • 发表日期：

    2025年

查看全文
  • Mauro Bruno,
  • Elena Catanese,
  • Erika Cerasti,
  • Massimo De Cubellis,
  • Fabrizio De Fausti,
  • Marco Di Zio,
  • Gerarda Grippo,
  • Giuseppe Lancioni,
  • Giulio Massacci,
  • Stefano Mugnoli,
  • Francesco Ortame,
  • Angela Pappagallo,
  • Francesco Pugliese,
  • Alessandra Righi,
  • Alberto Sabbi,
  • Francesco Sisti,
  • Donato Summa,
  • Luca Valentino

摘要

Over the past decade, National Italian Institute of Statistics (Istat) has recognized the potential of big data and machine learning (ML) in modernizing statistical production. This chapter outlines Istat’s journey from early experimentation to new statistical production pipelines based on Trusted Smart Statistics (TSS) principles. We discuss integrating big data sources—such as satellite imagery, web data, and sensor-based information—into official statistics and the challenges this paradigm shift poses. The chapter describes several research projects, including maritime traffic analysis using automatic identification system (AIS) data, urban green area estimation through remote sensing, and web intelligence for business statistics. Special attention is given to the methodological issues surrounding ML applications in official statistics, particularly in relation to estimation, imputation, and quality assessment. We examine the evolving role of data science skills within Istat, the institutional investments in human capital, and involvement in international research projects that have supported innovation. By integrating ML and big data into statistical processes, Istat aims to enhance data quality, improve timeliness, and provide more granular insights while ensuring compliance with statistical principles and data governance frameworks. The chapter concludes with considerations on future challenges and opportunities for official statistics in the datafied society.