Analysis of Sentimental on Twitter using Content Aware Support Vector Machines

In the era of the Internet’s rapid expansion, social networking platforms have emerged as a crucial medium for individuals to convey their emotions and opinions to a global audience. Text, images, audio, and video are all ways for people to express themselves. Twitter, a popular social network, allows users to tweet about their daily feelings and opinions. These platforms generate vast amounts of unstructured data continuously, necessitating prompt processing to gain insights into human psychology. This processing is accomplished through sentiment analysis, a technique that assesses the emotional polarity within texts, determining whether the author holds a -Ve, +Ve, or neutral stance toward a subject, service, individual, or place. This article investigates the use of four well-known data mining classifiers, specifically the decision tree, K-nearest neighbor, naive Bayes, and support vector machine, to analyze tweet sentiments. Our findings reveal that the support vector machine consistently yields superior results compared to other algorithms, demonstrating an improvement of 4.01% in accuracy for two-class datasets and 8.05% for three-class datasets.