Twitter Text Classification using Convolutional Neural Network Method

Text classification on social media platforms such as Twitter has become increasingly crucial. Convolutional Neural Networks (CNNs) have demonstrated their effectiveness across a range of natural language processing tasks, including text classification. The primary goal of this article is to create a reliable and precise text classification model for Twitter by employing CNNs. The CNN architecture is tailored for the text classification task through the application of one-dimensional convolutions on the word embedding. In this article utilize multiple convolutional layers with diverse kernel sizes to capture various levels of contextual information within the input text. Max-pooling layers are employed to extract the most pertinent features from the convolved results. To assess the performance of the text classification model based on CNNs, this study carry out experiments using a diverse dataset of Twitter messages. The dataset is annotated with various categories such as sentiment (positive, negative). Experimental results demonstrate that the proposed CNN model achieves competitive performance compared to state-of-the-art methods for text classification on Twitter. then compared the proposed model with some ML methods like Logistic Regression (LR), Naive Bayes (NB), Stochastic Gradient Descent (SGD), and k-nearest neighbours (KNN) and got the following accuracy: 98%, 97%, 89%, 83%, and 94%.