Utilization of Artificial Intelligence in Consumer Sentiment Analysis on Social Media to Support Marketing Strategy

Abstract

The rapid growth of social media platforms has transformed how consumers
express their opinions, making sentiment analysis a critical tool for
understanding consumer behavior. This research explores the use of Artificial
Intelligence (AI) in sentiment analysis, specifically through Natural Language
Processing (NLP) techniques, to analyze consumer sentiment on social media
platforms such as Twitter and Instagram. By employing sentiment
classification models, including BERT (Bidirectional Encoder
Representations from Transformers) and Logistic Regression with TF-IDF,
the study aims to uncover patterns in consumer sentiment and provide insights
to businesses for developing effective marketing strategies. The results
demonstrate that BERT outperforms Logistic Regression, offering higher
accuracy, precision, recall, and F1-score in sentiment classification.
Additionally, sentiment trend analysis highlights how consumer opinions
fluctuate over time in response to marketing campaigns, while sentiment
distribution analysis provides an overview of the general attitude toward
products. This study offers a comprehensive AI-driven framework for
businesses to improve customer satisfaction, optimize marketing efforts, and
enhance brand loyalty through real-time sentiment insights.