Modelling the Names of Jeffrey Beall's List of Possible Predatory Journals

On his web site, Jeffrey Beall proposes a list of possible predatory journals. The names of these journals are compared to the source publication list for the Web of Science of Thomson Reuter. By means of the lasso statistical method and logistic regression, we identify a list of words of titles that enables to discriminate between both lists. The titles of the possible predatory journals contains more often vague terms as `advanced', research', `international' or `sciences'. Some journal names are so characteristic that it is possible to almost predict with certainty that they are possible predatory ones.