Application of Identification in Literary Works Analysis Using Improved Random Forest

An identification of author is a significant characteristic of literary analysis, studied in Natural Language Processing (NLP). It supports to determine the most appropriate authors of the studies, comments from social media or news texts and so on. Thus, this research proposes the Improved Random Forest (IRF) approach for the Literary Work Analysis. This research introduced the author identification approach using Machine Learning (ML) approach, that is primarily categorized into various stages such as text pre-processing, feature extraction and classification. The stemming, lemmatization, tokenization are the methods used for the text pre-processing. Then, the Word 2 Vec is used for the feature extraction process. Finally, the IRF approach is utilized for the analysis of authors in literary work analysis. The experiments are employed on 1200 articles from 50 authors, which reaches a greater accuracy of 94.56% as compared to the traditional RF approach. The introduced approach for the literary work analysis outperformed the benchmark handcrafted features and established works on publicly available datasets.