An Automated Text Document Classification Framework using BERT

Due to the rapid advancement of technology, volume online text data from numerous various disciplines is increasing significantly over time. Therefore, more work needed create systems that can effectively classify in accordance with its content, facilitating processing and extraction crucial information. Since these non-automated use manual feature classification, which error-prone time-consuming by choosing best appropriate algorithms for traditional procedures are typically resource intensive (computational, human, etc.), not a viable solution. To address shortcomings approaches, we offer unique categorization strategy based on well-known DL algorithm called BERT. The proposed framework trained tested using cutting-edge datasets, such as UCI email dataset, includes spam non-spam emails, BBC News multiple categories tech, sports, politics, business, entertainment. system achieved highest accuracy 91.4% be used different organizations text-based high performance. effectiveness evaluated evaluation metrics Accuracy, Precision, Recall.