A Question Answering System For Interacting with SDMX Databases

Guillaume Thiry (1, 2) , Ioana Manolescu (1) , Leo Liberti (3)

1 CEDAR - Rich Data Analytics at Cloud Scale
2 X-DEP-INFO - Département d'informatique de l'École polytechnique
3 LIX - Laboratoire d'informatique de l'École polytechnique [Palaiseau]

Among existing sources of Open Data, statistical databases published by national and international organizations such as the International Monetary Fund, the United Nations, OECD etc. stand out for their high quality and valuable insights. However, technical means to interact easily with such sources are currently lacking. This article presents an effort to build an interactive Question Answering system for accessing statistical databases structured according to the SDMX (Statistical Data and Metadata Exchange) standard promoted by the abovementioned institutions. We describe the system architectures, its main technical choices, and present a preliminary evaluation. The system is available online.

Mots clés: Statistical databases; Question Answering; SDMX; data journalism; fact-checking

Domaines: Base de données [cs.DB]

Guillaume Thiry, Ioana Manolescu, Leo Liberti. A Question Answering System For Interacting with SDMX Databases. NLIWOD 2020 - 6th Natural Language Interfaces for the Web of Data / Workshop (in conjunction with ISWC), Nov 2020, Heraklion, Greece. ⟨hal-03021075⟩