The paper by Ouamour and Sayoud examines authorship attribution for a corpus of ancient Arabic travel writings (AAAT) by applying character‑based n‑gram features (unigrams to tetragrams) and four classifiers (Stamatatos distance, Manhattan distance, multilayer perceptron, and support vector machines), reporting up to 90 % correct attribution; it contributes empirical results to the fields of authorship attribution and stylometric text‑mining for Arabic language documents.
Bears on: authorship & stylometry · concepts & methods