Authorship Attribution in Bangla literature using Character-level CNN

Abstract

Characters are the smallest unit of text that can extract stylometric signals to determine the author of a text. In this paper, we investigate the effectiveness of character-level signals in Authorship Attribution of Bangla Literature and show that the results are promising but improvable. The time and memory efficiency of the proposed model is much higher than the word level counterparts but accuracy is 2-5% less than the best performing word-level models. Comparison of various word-based models is performed and shown that the proposed model performs increasingly better with larger datasets. We also analyze the effect of pre-training character embedding of diverse Bangla character set in authorship attribution. It is seen that the performance is improved by up to 10% on pre-training. We used 2 datasets from 6 to 14 authors, balancing them before training and compare the results.

Authorship Classification in a Resource Constraint Language Using Convolutional Neural Networks

“…They built a corpus consisting of 3125 passages and gained the highest accuracy (96%) with random forest than Naive Bayes (62%) and decision tree (85%) classifiers. Khatun et al [6] introduced a character-level CNN for attributing Bengali authorship. This system's performance decreased with an increased number of authors and sample texts.…”

“…The layer-wise weight values are stored through the metafile. To investigate the effect of authorship classification performance, LSTM [19], Char-level-CNN [6], SVM [10], SGD [50], Multilingual pre-trained BERT (M-BERT) [51] and Distil-BERT [52] classifiers are also implemented on the same datasets.…”

“…The feature vector is projected into an output block weight matrix with a shape of 18 × 384. The Soft-Max function produces AES using equation (6). The AES is defined as A = {as 1 , as 2 , as 3 , ..., as 18 } and finds the maximum expected value (Max(A)) with a corresponding index of the predicted author name.…”

“…In recent years, authorship classification has attracted much attention by Bengali Language Processing (BLP) re-There are several statistical and machine learning (ML) techniques that addressed the authorship attribution problem. The statistical methods are used for extracting stylometry features [2]- [5], character level embedding features [6], [7], and n-gram features [8], [9] for authorship attribution. Several ML techniques such as support vector machine (SVM), Naive Bayes (NB) [10]- [12], CNN [6], [13]- [15] are the most commonly used methods for authorship classification.…”

“…The statistical methods are used for extracting stylometry features [2]- [5], character level embedding features [6], [7], and n-gram features [8], [9] for authorship attribution. Several ML techniques such as support vector machine (SVM), Naive Bayes (NB) [10]- [12], CNN [6], [13]- [15] are the most commonly used methods for authorship classification. Nevertheless, stylometry, character level, and n-gram feature extractors cannot capture the sentence and document level semantic and syntactic features.…”