Brand Sentiment: Aspect-based Multilingual Sentiment Analysis for Strategic Brand Reputation Enhancement and Management in India

In the evolving digital landscape, brand reputation is shaped significantly by customer sentiment across diverse platforms. For businesses in India, where multilingual interactions are common, capturing sentiment in regional languages like Tamil and Kannada is crucial. This paper presents Brand Sentiment, an aspect-based sentiment analysis model designed to evaluate customer feedback and online interactions in both Tamil and Kannada. Our approach blends ML and NLP methods to identify sentiment polarity on specific brand aspects, offering insights for strategic reputation management. We employ a multilingual dataset and fine-tune transformer models for precise sentiment classification. The results demonstrate high accuracy and efficiency in identifying sentiment trends, allowing businesses to make informed decisions to enhance their brand’s reputation. This research fills a vacuum in regional language processing, which advances the rapidly expanding field of multilingual sentiment analysis., while also providing a scalable solution for Indian businesses.