Publication

Abstract : The high volume of user-generated content on digital platforms has highlighted the necessity of extracting meaningful insights from various languages. In sentiment analysis, identifying aspect terms is crucial for capturing the emotions of user opinions. This research paper introduces a specialized Aspect Term Extraction Technique for Hindi text, addressing the unique linguistic challenges posed by the language. The proposed method combines natural language processing (NLP) and deep learning techniques to automatically identify aspect terms from Hindi text data. Additionally, by integrating a lexicon-based approach to set the polarity of Hindi sentences and BERT multi-class classification for Aspect-Based Sentiment Analysis (ABSA), we achieved an accuracy of 86.97% in classifying 50K Hindi reviews.