schoonees / fakenewsthesis Goto Github PK
View Code? Open in Web Editor NEWThis project forked from franzschrepf/fakenewsthesis
For my master thesis at the Rotterdam School of Management, a data set of 369 news articles regarding the 2016 US presidential election was collected (199 real, 170 fake). Several text elements are extracted and the resulting model based on boosted classification trees is able to classify news content with an accuracy of roughly 91%.