The Research Department of Economics and Management of Tadbir Nikan Business, Marketing, and Finance Open In Press 2026 09 01 Sentiment Analysis of Customer Opinions in Iran: A Systematic Review Sentiment Analysis of Customer Opinions in Iran: A Systematic Review 1 15 EN Journal Article 2025 10 08 <p>The quick growth of digital commerce and online communication with clients has generated enormous volumes of user-produced textual data, requiring automation to extract insight into consumer opinion. This systematic review explains the contemporary status of sentiment analysis studies about Iranian consumers' opinions, focusing on datasets, techniques, evaluation procedures, and limitations contained in the literature. According to PRISMA 2020 guidelines, pertinent studies were located in both the IEEE Xplore and MagIran databases by both English and Persian search terms. The studies selected for inclusion were evaluated according to a modified JBI checklist and separated according to their employed methodological approach, including lexicon-based rules, pattern-mining and graph-based methods, neural networks, and transformer-based pre-trained models. The results show that although the transformer model and hybrid methods utilizing deep learning are more efficient than the traditional techniques in use, both the results of the transformer and hybrid methods are restricted by limited and domain-specific datasets in Persian, various methods of preprocessing, and the absence of standard evaluation. Linguistic issues, including extensive morphology, different representations of orthographical spelling, and successive code-mixing of different languages, also contribute to the difficulties of adequately interpreting sentiment issues with accuracy. The analysis exposes several major research deficiencies, necessitating the generation of large, standardized, and multi-domain corpora in Persian suitable for inter-application evaluation of sentiment analysis procedures. In addition, sentiment analysis procedures suitable to the linguistic demands of Persian, called for in adopting extensive and standard datasets, are suggested, which will provide easier explainable and reproducibility of sentiment analysis.</p> https://www.bmfopen.com/index.php/bmfopen/article/download/385/289