Author Identification using Stylometric Features
Daniel Pavelec, Edson Jose Rodrigues Justino, Luiz S. Oliveira · INTELIGENCIA ARTIFICIAL · 2007
"In this work we present a strategy for author identification for documents written in Portuguese. It takes intoaccount a writer-independent model which reduces the pattern recognition problem to a single model andtwo classes, hence, makes it possible to build robust system even when few genuine samples per writer areavailable. We also introduce a stylometric feature set, which is based on the conjunctions of the Portugueselanguage. Experiments on a database composed of short articles from 10 different authors and Support VectorMachine (SVM) as classifier demonstrate that the proposed strategy can produced results comparable to theliterature"