Chakraborty (2013) presents a comparative study of authorship identification for Bengali texts, employing fine‑grained stylistic features within both a statistical similarity framework (cosine similarity, chi‑square, Euclidean distance) and machine‑learning classifiers (Decision Tree, Neural Network, SVM), and evaluates their performance via 10‑fold cross‑validation. The work contributes to stylometry‑based authorship attribution and demonstrates the relative importance of individual linguistic markers across models.
Bears on: authorship & stylometry · concepts & methods