Abstract
Authorship classification is a method of automatically determining the appropriate author of an unknown linguistic text. Although research on authorship classification has significantly progressed in high-resource languages, it is at a primitive stage in the realm of resource-constraint languages like Bengali. This paper presents an authorship classification approach made of Convolution Neural Networks (CNN) comprising four modules: embedding model generation, feature representation, classifier training and classifier testing. For this purpose, this work develops a new embedding corpus (named WEC) and a Bengali authorship classification corpus (called BACC-18), which are more robust in terms of authors’ classes and unique words. Using three text embedding techniques (Word2Vec, GloVe and FastText) and combinations of different hyperparameters, 90 embedding models are created in this study. All the embedding models are assessed by intrinsic evaluators and those selected are the 9 best performing models out of 90 for the authorship classification. In total 36 classification models, including four classification models (CNN, LSTM, SVM, SGD) and three embedding techniques with 100, 200 and 250 embedding dimensions, are trained with optimized hyperparameters and tested on three benchmark datasets (BACC-18, BAAD16 and LD). Among the models, the optimized CNN with GloVe model achieved the highest classification accuracies of 93.45%, 95.02%, and 98.67% for the datasets BACC-18, BAAD16, and LD, respectively.
Highlights
Authorship classification is a long-established research topic in Natural Language Processing (NLP) that deals with the difficulty of identifying the author against a particular text
Based on the intrinsic evaluation performance, a total of 9 top-performing embedding models are selected for the authorship classification task
Three models are chosen from Global Vectors for Word Representation (GloVe), three from FastText and three from Word2Vec embeddings based on the highest Pearson and Spearman correlation scores
Summary
Authorship classification is a long-established research topic in Natural Language Processing (NLP) that deals with the difficulty of identifying the author against a particular text. Other authors may prefer to apply particular clauses, specific tense, distinguished sentence structure or open and close sentences with an appropriate grammatical constituent These features can be used in identifying the authorship of a particular writing. Authorship classification is a well-established research topic for high resource languages (e.g., English and other European languages) due to the availability of authorship corpus, feature extractors and classification techniques. It is a challenging task for a low-resource language like Bengali due to the shortage of linguistic resources and techniques [1].
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have