Current problems of forensic authorship analysis and the possibility of their solution with the use of computer methods: problems and prospects

Active development of Internet communication in recent years caused an increase in the number of forensic text examinations aimed at identifying and profiling (i.e. inferring gender, age, personality, etc. of the author from textual analysis) the author of written texts. Despite the availability of proven methodological recommendations for the production of such examinations, in this area there are many unresolved problems associated mainly with the emergence of new research objects. In addition, the existing expert practice does not fully utilize the achievements of corpus, computer, and quantitative linguistics. In this situation, there is a gap between the “qualitative” and “quantitative” methods of textual authorship analysis, which hinders further development of both theoretical research in the area of authorship attribution and profiling and an increase in the level of objectivity and reproducibility of forensic authorship analysis. The paper represents some typical tasks solved by a forensic expert; describes the characteristics of the objects of forensic authorship analysis, and determines the main difficulties forensic experts face in the course of this analysis. The possibilities of using existing computer methods to solve these tasks are analyzed. It is shown that not all the existing computer methods are useful for forensic authorship analysis. We also highlight the ways of development of forensic authorship analysis related to further theoretical research in the field of idiolect using corpus data and natural language processing techniques.