Scholarlink

Sentiment analysis, a crucial task in Natural Language Processing (NLP), aims to extract and classify opinions, emotions, and attitudes expressed in text. Analyzing customer reviews can provide valuable insights into service quality and customer satisfaction in the context of courier services. However, manual analysis of large volumes of unstructured text is time-consuming and prone to human error. This paper proposes a CNN-based sentiment analysis system to address these challenges. CNNs are well-suited for processing sequential data, such as text, and have demonstrated superior performance in various NLP tasks. Our proposed methodology involves applying CNNs to capture and analyze hierarchical representations of text data. The model employs convolutional layers for feature detection, pooling layers for dimensionality reduction, and fully connected layers for final classification decisions. The model is trained on a dataset of courier service reviews to detect and classify sentiments as positive or negative automatically. Our results demonstrate an impressive overall accuracy of 88.69%, highlighting the model’s effectiveness in accurately distinguishing between sentiments.

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

Sentiment Analysis on Courier Services in Malaysia Using Convolutional Neural Network (CNN) Algorithm

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Ummu Fatihah Mohd Bahrin,
  • Ezdiyad Effendy,
  • Siti ‘Aisyah Sa’dan,
  • Normalina Ibrahim

摘要

Sentiment analysis, a crucial task in Natural Language Processing (NLP), aims to extract and classify opinions, emotions, and attitudes expressed in text. Analyzing customer reviews can provide valuable insights into service quality and customer satisfaction in the context of courier services. However, manual analysis of large volumes of unstructured text is time-consuming and prone to human error. This paper proposes a CNN-based sentiment analysis system to address these challenges. CNNs are well-suited for processing sequential data, such as text, and have demonstrated superior performance in various NLP tasks. Our proposed methodology involves applying CNNs to capture and analyze hierarchical representations of text data. The model employs convolutional layers for feature detection, pooling layers for dimensionality reduction, and fully connected layers for final classification decisions. The model is trained on a dataset of courier service reviews to detect and classify sentiments as positive or negative automatically. Our results demonstrate an impressive overall accuracy of 88.69%, highlighting the model’s effectiveness in accurately distinguishing between sentiments.