Authors: Dan Gillman (BLS), Franck Cotton (INSEE), Yves Jaques (FAO) Title: eXtended Knowledge Organization System (XKOS) Abstract: XKOS is a family of Resource Description Framework (RDF) classes and properties that extend the expressive power of SKOS (Simple Knowledge Organization System) to meet the specific modeling requirements of statistical classifications. SKOS was published in 2009 as a World Wide Web Consortium recommendation designed for use within the Linked Open Data (LOD) community. It provides a framework for representing knowledge organization systems, including classifications, thesauri, code sets, taxonomies and similar resources. As statistical agencies are beginning to disseminate data using LOD approaches, the suitability of SKOS for publishing statistical classifications required review. The SKOS specification is based on the now withdrawn standard ISO 2788 - Guidelines for the establishment and development of monolingual thesauri. This standard and the SKOS specification derived from it have limited semantics for expressing relationships between concepts and no levels within which to organize classification systems. At workshops in Dagstuhl, Germany in 2011 and 2012 on Semantic Statistics for Social, Behavioural, and Economic Sciences: Leveraging the DDI Model for the Web, a group was formed to address the issue of how to extend SKOS to meet the needs of statistical organizations. A number of extensions were deemed important enough for inclusion, including the objects needed to transpose the Neuchâtel model and more refined semantic relations between concepts, and the results of that work are reported here.