Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{205433,
        author = {P. Anbumani and Dr. K. Selvaraj},
        title = {AMAZON PRODUCT REVIEW-BASED PRODUCT RECOMMENDATION USING OPTIMIZED BERT APPROACH},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {6515-6524},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=205433},
        abstract = {As Amazon e-commerce expands rapidly, a significant number of consumers are posting reviews of products online. The business-to-consumer relationship, as reflected in product reviews, is essential for understanding consumers' sentiments about services and products. Education on online shopping has led to an increase in online consumer reviews, resulting in a more intricate process of product selection. In the past, sentiment analysis (SA) frequently failed to recognize the emotions conveyed in the posts on the product review. This paper offers a Particle Swarm Optimization with a Bidirectional Encoder Representation by the Transformers (PSO-BERT) to address the above issues. The information gathered in this proposed method will include customer reviews on different products based on the Amazon data. Prior to calculating, the acquired data were filtered with the SentiBagofWordNet (SBWNet) algorithm to eliminate irrelevant text content (reviews) and keep only the required data. Following that, the Term Frequency Word2vector (TFW2v) technique is used to determine the relationship between words in reviews. In this paper, the Random Gradient Emoji Weight Rate (RGEWR) method is proposed for identifying text-weighted emojis in the Amazon products dataset. The proposed PSO-BERT method is expected to be beneficial for product ranking, leveraging a bio-inspired, optimized BERT approach. The findings indicate that the suggested PSO-BERT demonstrates the best training and testing precision on a dataset gathered on Amazon.},
        keywords = {Emoji weigh identification, bio inspired, product recommendation, ranking, sentiment analysis, reviews.},
        month = {June},
        }