Demographic and Social Statistics

This webinar serves as the inaugural session of a specialized series launched by the United Nations Statistics Division (UNSD), aimed at bridging the gap between high-level international recommendations and the practical, on-the-ground implementation of modern advancements in Artificial Intelligence (AI) and Machine Learning (ML). The session explores how National Statistical Offices (NSOs) are utilizing these technologies to drive efficiency, enhance data quality, and improve user engagement across the census cycle. Moving beyond theoretical discussion, the programme focuses on three core operational transformations: leveraging ML and GeoAI for automated building footprint detection and boundary delineation in census mapping; utilizing Natural Language Processing (NLP) and historical data models to automate the coding of complex, open-ended responses like industry and occupation; and deploying public-facing chatbots, virtual assistants, and NLP-driven dissemination tools to simplify data exploration and support field staff.

• Concept Note
• Video recording

• Opening and welcoming remarks
  • - Francesca Grum, Assistant Director, Chief, Social Statistics Branch, UNSD
• Presentations
• a. AI for Census Mapping: Opportunities and Challenges in Practice (pdf)
• - by Florian Nika, Geospatial Software Engineer, Research and Development Unit of the Italian Geographical Society
• b. Automating open-response coding using AI in Chile's 2024 Population and Housing Census (pdf)
• - by Edgardo Cerda, Senior Census Analyst, National Institute of Statistics of Chile
• c. AI Applications in Census Operations and Data Dissemination (pdf)
• - by Sally Lim, Deputy Director, Household Surveys and Expenditure, and Yi Xin Chua, Manager, Communications and Engagement, Singapore Department of Statistics
• Q&A
• Closing remarks