A Framework for the Authorship Identification in Research Papers

Abstract

Authorship identification and inherent plagiarism detection are crucial to academic and literary ethics. Traditional EPD techniques compare papers to digitalized or internet-available sources, missing plagiarized content from novels or textbooks. Adding non-contributors' names to papers is unethical and undermines motivated researchers' reputation. This study uses stylometric traits to determine authorship and plagiarism without external sources. Stylometric indicators including writing style, language, and sentence structure are used to assign authors to document parts and uncover discrepancies that indicate numerous contributors. Clustering is used to count the authors in a manuscript, solving unethical authorship attributions and concealed plagiarism.
The study analyzes methods, identifies limits, and recommends anomaly detection and text feature improvements. The findings show that the suggested method can detect multi-author contributions and non-digital plagiarism. This study provides a complete authorship identification and intrinsic plagiarism detection method to promote academic integrity, discourage unethical activities, and inspire real researchers.