Automatic Identification of Authors’ Writing Style with Computer-Based Stylometry Methods: A Case Study on Indonesian Literary Texts

The study of authors’ writing styles has long been a significant area of research in literary studies and forensic linguistics. In recent years, the advancement of computational methods has led to the development of automatic stylometry techniques, which can identify and analyze unique features of an author’s writing style. This research aims to apply computer-based stylometry methods to the identification of authors’ writing styles in Indonesian literary texts, a field that has not been extensively explored. By leveraging these methods, the study seeks to explore the potential of automated tools in distinguishing between authors based on their stylistic markers, such as vocabulary choice, sentence structure, and punctuation usage. The research utilizes stylometry techniques, including machine learning algorithms and text mining, to analyze a corpus of Indonesian literary texts. These methods allow for the extraction of distinct stylistic features from the texts, which are then used to classify and differentiate between authors. The results show that machine learning models, particularly those utilizing n-grams and frequency analysis, can accurately distinguish between authors with high precision. The study concludes that computer-based stylometry offers a promising approach for analyzing Indonesian literary texts and can be used effectively in authorship attribution. This method provides valuable insights into the unique stylistic signatures of authors and can aid in the study of literary history and forensic analysis.