Aspect-Aware Sentiment Analysis of Code-Mixed Amazon Indian Reviews Using Roberta
International Journal of Creative and Open Research in Engineering and Management · 25.06.2026 · 02 (06) · 1-5
DOI: 10.55041/ijcope.v2i6.320 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Creative and Open Research in Engineering and Management, online erschienen am 25.06.2026. Verfasst von Prof. D. V. Mehta Prof. D. V. Mehta und 4 weiteren Autor:innen.
Abstract (Originalsprache)
With the increasing adoption of online shopping sites in India, there is now a huge amount of review generation in Indian code-switched languages like Hinglish, Marathi-English, etc. Conventional sentiment analysis tools face challenges in analyzing such multilingual data with errors in grammatical structures, vocabulary, and context understanding. The paper proposes an aspect-based sentiment analysis tool that uses an XLM-RoBERTa model in a Django web application for code-mixed Amazon product reviews in India. Reviews in the database are dynamically fetched using the Rainforest API with the help of ASIN numbers. This classifier categorizes the general sentiment as being positive, negative, or neutral while simultaneously extracting sentiments with respect to features like battery life, camera quality, price, and performance, etc. Experimentation with met-rics such as precision, recall, and F1-Score indicates better context comprehension as compared to conventional polarity based systems. These aspects allow for better interpretation for consumers, sellers, and researchers alike. Index Terms—Code-Mixed Language, Hinglish, Aspect-Based Sentiment Analysis, XLM-RoBERTa, Multilingual NLP, E-commerce Analytics
Zitieren
APA
Mehta, P. D. V. M. P. D. V., Ranaware, S. R. S., Kharat, R. K. R., Bargaje, K. B. K., & Khude, N. K. N. (2026). Aspect-Aware Sentiment Analysis of Code-Mixed Amazon Indian Reviews Using Roberta. International Journal of Creative and Open Research in Engineering and Management, 02(06), 1-5. https://doi.org/10.55041/ijcope.v2i6.320
Vancouver
Mehta PDVMPDV, Ranaware SRS, Kharat RKR, Bargaje KBK, Khude NKN. Aspect-Aware Sentiment Analysis of Code-Mixed Amazon Indian Reviews Using Roberta. International Journal of Creative and Open Research in Engineering and Management. 2026;02(06):1-5. doi:10.55041/ijcope.v2i6.320
BibTeX
@article{Mehta2026,
author = {Mehta, Prof. D. V. Mehta Prof. D. V. and Ranaware, Shriraj Ranaware Shriraj and Kharat, Rutuja Kharat Rutuja and Bargaje, Karishma Bargaje Karishma and Khude, Nikita Khude Nikita},
title = {Aspect-Aware Sentiment Analysis of Code-Mixed Amazon Indian Reviews Using Roberta},
journal = {International Journal of Creative and Open Research in Engineering and Management},
year = {2026},
volume = {02},
number = {06},
pages = {1--5},
doi = {10.55041/ijcope.v2i6.320}
}
EUVIMED