Sentiment Analysis of Skincare Product Reviews in Indonesian Language using IndoBERT and LSTM
Jessica Hidayat Computer, Shilvia Meidhi Honova, Vianny Pangesa Computer, Christiella Abinosy Setiawan, Ivan Halim Parmonangan, Diana Diana
Abstract
This research focuses on analyzing sentiment to- wards skincare products in the Indonesian language, employing the IndoBERT approach and LSTM classification. The study aims to provide consumers with accurate information before purchasing skincare products and assist companies in improving product quality based on customer feedback. The method entails feature extraction with IndoBERT and classification using LSTM. The evaluation demonstrates excellent performance, with accu- racy reaching a value of 92.6%, indicating the effectiveness of the IndoBERT and LSTM approaches in sentiment classification for Indonesian skincare product reviews. These results lay a strong foundation for further development in sentiment analysis and contribute to the advancement of natural language processing technology in the context of skincare products in the Indonesian language.