The Key Factors and Their Influence in Authorship Attribution

Abstract

Authorship attribution has a long history started since 19th century. Existing studies have used different sets of stylometric features and computational methodologies on a variety of corpus with different lengths and genres. This study presents a protocol to perform a systematic literature review (SLR) to identify the best combination of stylomet-ric features and computational methodology. Specifically, we formulate an SLR protocol that can be used to conduct a literature survey to help answer like (i) whether it is possible to identify the authorial style of an author regardless the genre and length of the text, and (ii) how to select specific stylometric features and computational methodology. We also conduct an example of how the proposed SLR protocol can be used as a template for publication extraction and filtering for an SLR on authorship attribution.