Classification des systèmes de monnaies non-bancaires : ce que disent les données du Web

This article proposes a classification of the projects of non-banking currencies based on a lexical analysis on the Web data. The advantage of this method is to offer an endogenous typology of the initiatives, based on the way the various currencies are presented on the internet. She so allows to by-pass the difficulty met in the literature which faces troubles in bringing to the foreground a clear taxonomy from exogeneous elements. The corpus was created from 32 keywords referring to the complementary currencies. The first 10 Urls of results under the search engine Google for each of the keywords was selected, then their text contents extracted. The corpus is so constituted by 320 web pages, corresponding to 1210 pages of text and 342 585 words, that is 17 939 segments of 20 successive occurrences. A classification of the corpus by a downward hierarchical analysis endogenously creates five different classes, allowing us to operate distinctions not only between the projects of non-banking currencies, but also between these and the standard monetary system. An analysis of similarity concludes that all the currencies define themselves with regard to the conventional currency, except the local exchange trading systems (LETs).