Scholarlink

The e-commerce age allows consumers to purchase anything they desire. With the help of the internet, consumers can easily view and purchase any product. When making an online purchase, the quality of the product is crucial. Sentiment analysis is a useful tool for identifying the emotions present in a text and classifying reviews into three categories: positive, negative, and neutral. Despite the fact that several prior studies have been done to examine customer evaluations, we have always been at a loss as to how best to examine the vast quantity of feedback that customers have to offer. Many models, including CNN, SVM, LSTM, Bi-LSTM, CNN-LSTM, and Decision Tree, are presented in this paper. With an accuracy of 98%, CNN-Bi-LSTM outperformed the others on the dataset. In addition, for comparative analysis, we have employed other word embedding methods, including TF-IDF, Word2Vec, and GLOVE. The experiment demonstrates that this model outperforms the other models in terms of precision, recall, and F1-Score. This is a useful technique for extracting sentiment from customer feedback.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An Evaluation of Advanced Neural Network and Machine Learning Models for Sentiment Analysis on Product Reviews

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Md. Emon,
  • Niloy Roy,
  • Tahmina Akter,
  • Mst. Sharmin Akter,
  • Tanjim Mahmud,
  • Nurullaeva Nodira,
  • Atayev Shokir,
  • Abubokor Hanip,
  • Mohammad Shahadat Hossain

摘要

The e-commerce age allows consumers to purchase anything they desire. With the help of the internet, consumers can easily view and purchase any product. When making an online purchase, the quality of the product is crucial. Sentiment analysis is a useful tool for identifying the emotions present in a text and classifying reviews into three categories: positive, negative, and neutral. Despite the fact that several prior studies have been done to examine customer evaluations, we have always been at a loss as to how best to examine the vast quantity of feedback that customers have to offer. Many models, including CNN, SVM, LSTM, Bi-LSTM, CNN-LSTM, and Decision Tree, are presented in this paper. With an accuracy of 98%, CNN-Bi-LSTM outperformed the others on the dataset. In addition, for comparative analysis, we have employed other word embedding methods, including TF-IDF, Word2Vec, and GLOVE. The experiment demonstrates that this model outperforms the other models in terms of precision, recall, and F1-Score. This is a useful technique for extracting sentiment from customer feedback.