Identification of Disputed Writings in Tamil Articles Using Multivariate Statistical Techniques
G. Manimannan, R. Lakshmi Priya · IOSR Journal of Mathematics · 2014
Many interesting problems are associated with the science of authorship attribution andStylometry Analysis.By quantifying relevant features related to literary style, it is possible to classify articles written by different writers and also attribute authorship to newly discovered texts.Literary style attracts the opportunity to introduce and utilise many classical multivariate statistical techniques.In this paper, an attempt is made to attribute authorship on the basis of stylistic features of certain articles written on Indian freedom movements published in the magazine India.Application of Multivariate Hotelling 2 T and Cluster Analysis shows that the unknown articles are significantly similar of known articles.This indicates that the Stylometry evidence do place the unknown articles are attributed to known articles.The control article is isolated from known and Unknown articles.Hence the two unattributed articles can be associated with the writings of Bharathiar.The writing style of other authors is also extracted in this study.