Sentiment Analysis of Computer Mediated Communication in Social using Natural Language Processing

Abstract

The massive interactions on social mediaplat forms had created a luxury of computer-mediated communication (CMC) languages in recent times, especially on the X (formerly Twitter) Platform. Resources required in extracting and analyzing these enormous expressions whether for public perception, market trends, or social dynamics are incredibly huge and can also be complex to handle. The comparative investigation of the CMC based on the accuracy of the interpreted sentiments is expressed within the Google Natural Language Processing(NLP) API model. The results of experts� analysis with that of the Google NLP model using sizable data of CMC from X were compared. The X comments on the declaration of the state of emergency by the Nigeria President-Bola Ahmed Tinubu- in Rivers State on the 18^th of March 2025 as posted by its handlers were the subjects of analysis. Identification and categorization of sentiment polarity whether positive, negative, or neutral were carried out by the model. Indices such as linguistic variations, context-dependent sentiment, sarcasm, and irony were used in order to understand the influence on the accuracy and reliability of sentiment analysis results of the tool. The outcome of this paper reveals the remarkable strengths and weaknesses of an Google NLP model in analyzing sentiment present in the CMC in social media platforms.