Protocol
Always starting from datasets gained in protocol [2.1] (text bodies, fields, authors, years), first of all we made a list of authors who wrote more articles and papers. Here we did not consider website authors because they are not relevant: expert journalists or academic researchers are the only cited in wherever. We took the first 56 authors from this list (those who wrote a considerable number of documents) and calculated how many times they were cited in the texts among all our three fields. To do this, before, we needed to have the authors written in the same manner in the list of 56 and in the text bodies, too: the solution was to make a Python script able to convert names and surnames of authors in the list and in the bodies, from whatever format to a unique format: namesurname. Once achieved this result, we could use an R script (using “tm” library) to count how many times each author has been cited within LexisNexis, Scholar and Google search text bodies.