LAUSR

Authorship attribution (AA) is an important task, as it identifies the author of a written text from a set of suspect authors. Different methodologies of anonymous writing have been discovered with the rising usage of social media. This anonymous writing leads to an increase in malicious and suspicious activities, and anonymity makes it difficult to find the suspect. AA helps to find the writer of a suspect text from a set of suspects. Different social media platforms, such as Twitter, Facebook, and Instagram, are used regularly by the users for sharing their daily life activities. Finding the writer of microtexts is considered the toughest task due to the shorter length of the suspect piece of text. We present a Capsule-based convolutional neural network (CNN) model over character $n$ -grams for performing the AA task. Capsule with kervolutional neural networks (KNNs) has also been utilized for this task. We also present different analyses of our developed system, which improves the interpretability of our developed system. Heat-maps for different models illustrate the relevant text fragments for the prediction task. A standard Twitter data set is used for evaluating the performance of the developed systems. The experimental evaluation shows that capsule-based CNNs and capsule-based KNNs perform competitively and are able to outperform previous methods. The source codes and the supplementary file are available here https://github.com/chanchalIITP/AuthorIdentification.