Authorship Semantical Identification Using Holomorphic Chebyshev Projectors

Abstract

Text attribution and classification, for both information retrieval and analysis, have become one of the main issues in the matter of security, trust and copyright preservation. This paper proposes an innovative approach for text classification using Chebyshev polynomials and holomorphic transforms of the coefficients space. The main advantage of this choice lies in the generality and robustness of the proposed semantical identifier, which can be applied to various contexts and lexical domains without any modification.