Intrinsic Detection of Plagiarism based on Writing Style Grouping

In this paper, we tackle the task of intrinsic plagiarism detection, also referred to as author diarization. This task deals with identifying segments within a document written by multiple authors [2]. The main goal is to discover deviations in the writing style, looking for parts of the document that could potentially be written by another person [4]. In this paper, we present our hybrid approach that constructs a style function from stylometric features and detects the outliers. The proposed approach has been evaluated on two publicly available corpora. The obtained results outperform the ones obtained by the best state-of-the-art methods.