Using Wikidata and Wikipedia for assisted generation of a structured multilingual vocabulary about the Covid-19 pandemic

Abstract

A method for quickly and dynamically building controlled vocabularies, especially for the media, using Wikidata and Wikipedia as sources of terminological information, is proposed. The method is applied to construct a vocabulary about the Covid-19 pandemic. For this purpose, it is proposed to exploit the structure of items and properties of Wikidata and links and backlinks of Wikipedia articles. Using a process based on the definition of Wikidata relationship expansion rules, an algorithm was designed, starting from a set of initial items and then being executed in successive iterations, followed by a review of the results. In this way, the Wikidata entities relevant to the thematic coverage of the vocabulary are collected. The algorithm has been implemented in an open-source application whose results for the Covid-19 pandemic vocabulary collection have been published in a repository. The algorithm can be used to verify the results using the same or other expansion rules or applied to compile vocabularies in other thematic areas. The results in terms of the elements collected in each iteration and the validation proposal through the links and backlinks of Wikipedia articles are also analyzed. The application of SKOS to achieve an interoperable representation of vocabularies obtained by this method is proposed as future work.