Abstract: Text conversion and categorization are fundamental machine learning operations in which an object is assigned to a subset of unnamed candidates. We propose a novel text classification approach based on Deep Learning in this research (DL). Our suggested method has a lot of appealing features: it gathers certain metadata from each item and builds the training set train first. The weight achieved under each class mark was used to classify each object. The Recurrent Neural Network (RNN) was employed in the proposed classification model. For training and testing, the streaming text was dispersed using cross-fold validation. The 70-30% of the data was used for training and research purposes, respectively. The consequences of the strategy in the chapter on outcomes are revealed through a partial implementation. The proposed method outperforms traditional approaches to text classification, according to the results of the experimental study.

Keywords: Text processing, Machine learning, deep learning, natural language processing, data analysis, artificial intelligence

PDF | DOI: 10.17148/IJARCCE.2021.105170

Open chat
Chat with IJARCCE