%0 Journal Article
%T Principal Component Analysis and Neural Networks for Authorship Attribution
%A Mehmet Can
%J Southeast Europe Journal of Soft Computing
%D 2012
%I 
%X A common problem in statistical pattern recognition isthat of feature selection or feature extraction. Feature selectionrefers to a process whereby a data space is transformed into a featurespace that, in theory, has exactly the same dimension as the originaldata space. However, the transformation is designed in such a waythat the data set may be represented by a reduced number of"effective" features and yet retain most of the intrinsic informationcontent of the data; in other words, the data set undergoes adimensionality reduction.In this paper the data collected by counting selected syntacticcharacteristics in around a thousand paragraphs of each of thesample books underwent a principal component analysis performedusing neural networks. Then, first of the principal components areused to distinguish authors of the texts by the use of multilayerpreceptor type artificial neural networks.
%K principal components
%K authorship attribution
%K stylometry
%K text categorization
%K function words
%K stylistic features
%K syntactic characteristics
%K multilayer preceptor
%K artificial neural network
%U http://www.scjournal.com.ba/index.php/scjournal/article/view/18/14