Special Track 10

Harnessing AI and data science to strengthen official statistics.

From the “lab” to the factory

Special Track Chairs:

Description

Official statistics is a key public infrastructure to deliver fit for purpose quality information for decisions to different users, including citizens, business and governments.

In recent years, a mismatch has emerged between the public demand for timely and detailed data to inform decisions and the traditional supply of trustworthy statistics.

To address this challenge statistical offices have invested in data innovation, harnessing the power of digital data and technologies supported by partnerships with academia and private sector. This has led to an acceleration of the use of AI and data science technologies in statistics that is rapidly moving from experimental projects to production-ready operations.

The introduction of new technologies and the use of new data sources, frequently in combination with traditional ones such as sample survey, census data and administrative records, has highlighted the importance of developing new frameworks for data integration and the production of high quality official statistics in the digital age.

These include methodology, ethics and quality standards, operational models, transparency and explainability frameworks for data integration, human centred AI and data science digital solutions, partnerships with stakeholders in the data ecosystem, secure technology infrastructure platforms and privacy enhancing technologies.

Submissions to the call should relate to conditions and mechanisms for moving from experimentation to high-quality data dissemination, the issue of data integration to produce new insights, the development of methodological, ethics and quality frameworks, the use of explainable AI in data analysis, processing and production of relevant insights, governance models enabling collaboration between government, academia and industry in the context of producing sustainable data for official statistics, the development and implementation of privacy, confidentiality and data security mechanisms protecting individuals and businesses, and the creation of sustainable partnerships between official statisticians, governments, academia and industry.

The special track session will solicit contributions from different stakeholders to explore how the official statistics system is addressing these challenges, both at national and international level and what new governance models are emerging to ensure a sustainable and human-centric use of AI and data science for official statistics.