Sentiment Analysis Using Large Language Models : A Review

This review examined sentiment analysis using Large Language Models and summarized research involving multilingual classification, aspect-based analysis, social-media sentiment, user reviews, and comparisons between LLMs and traditional machine-learning approaches. The literature shows that LLMs offer flexible prompt-based capabilities, but their performance depends on the task, domain, language, prompting strategy, and evaluation setting. The review also highlights the limitation of binary sentiment analysis for applications where neutral opinions and subtle dissatisfaction are important. A five-class framework can provide a more detailed representation of user feedback. For the proposed Google Gemini review-analysis study, TF-IDF and machine-learning classifiers provide a clear baseline, while Accuracy, Precision, Recall, and F1-score provide standard evaluation measures. Future research can extend this work through larger datasets, multilingual and aspect-based analysis, LLM-versus-ML comparisons, explainability, and real-time monitoring. These directions can support the development of more systematic and reliable sentiment-analysis systems for generative-AI applications.

[1]     Krugmann, Jan Ole, and Jochen Hartmann. "Sentiment analysis in the age of generative AI." Customer Needs and Solutions 11.1 (2024): 3. [2]     Šmíd, Jakub, Pavel Přibáň, and Pavel Kral. "LLaMA-based models for aspect-based sentiment analysis." Proceedings of the 14th workshop on computational approaches to subjectivity, sentiment, & social media analysis. 2024. [3]     Wojdecka, Agnieszka, Jakub Gromadziński, and Krzysztof Walczak. "Effectiveness of large language models and software libraries in sentiment analysis." Applied Computer Science 21.2 (2025). [4]     Jin, J. "Harnessing the Power of Large Language Models for Real-World Sentiment Classification on Social Media." ADVANCES IN ECONOMICS, MANAGEMENT  AND  POLITICAL  SCIENCES Учредители: Eliwise Academy Ltd 156.1 (2025): 214- 224. [5]     Bojić, Ljubiša, et al. "Comparing large Language models and human annotators in latent content analysis of sentiment, political leaning, emotional intensity and sarcasm." Scientific reports 15.1 (2025): 11477. [6]     Alabduljabbar, Reham. "User-centric AI: evaluating the usability of generative AI applications through user reviews on app stores." PeerJ Computer Science 10 (2024): e2421. [7]      Demirbilek, Mustafa, and Sevim Özulukale Demirbilek. "Restaurant Review Sentiment and SWOT Analysis: using AWS and GPT-4 Large Language Models." Acta Infologica 9.2 (2025): 491-511. [8]      Niimi, Junichiro. "Dynamic sentiment analysis with local large language models using majority voting: A study on factors affecting restaurant evaluation." arXiv preprint arXiv:2407.13069 (2024). [9]      Wang, Zengzhi, et al. "Is ChatGPT a good sentiment analyzer? A preliminary study." arXiv preprint arXiv:2304.04339 (2023). [10]  Zhang, Wenxuan, et al. "Sentiment analysis in the era of large language models: A reality check." Findings of the association for computational linguistics: NAACL 2024. 2024.