Building taxonomies on the base of vocabulary in real use in the system, Leopoldina.pl use case

Taxonomies are attempts of ordering world in semantic field. They are also a useful tool of knowledge organization, facilitating domain information retrieval. Taxonomy or thesauri creation requires extensive and thorough analytical process, engaging information and knowledge from variety of sources. Typically it is a longterm activity. Such traditional approach failed for knowledge organization gathered on Leopoldina.pl platform. The aim of the service is to present heterogenic and differentiated digital resources of University of Wrocław (UWr). Due to lack of success of creation an universal taxonomy in given period of time, an bottom-up approach was proposed. The entire range of vocabulary indexed in Leopoldina.pl platform was divided into discipline categories which were used as the base for taxonomies and thesauri creation. In current paper we describe the bottom-up process of taxonomy creation. Programming tools (in Python) used for domain dictionaries creation are presented. We provided also evaluation of external organized knowledge sources (like Wikipedia, GBIF ̶ the Global Biodiversity Information Facility and other domain-specific thematic portals) for automatic knowledge handling and taxonomies creation.