Mathematical World Knowledge Contained in the Multilingual Wikipedia Project

Abstract

The purpose of this project is to test and evaluate an approach for Formula Concept Discovery (FCD). FCD aims at retrieving a formula concept (in the form of a Wikidata item) together with its defining formula within documents, in this case 100 English Wikipedia articles. To correctly identify the defining formula of a Wikipedia article, this approach searches for shared formulae across Wikipedia articles available in different languages. The formula shared in the most languages is then assumed to be the defining formula. The results show that neither this approach alone nor a combination with an existing approach that considers the order of the formulae inside an article leads to satisfying results. It is thus concluded that the number of times a formula is shared across a Wikipedia article in different languages is not a good indicator to determine the defining formula with the current approach. Consequently, several ideas for further research are proposed which could improve the results.