Article: Deep learning semantic understanding and classification of student online public opinion for new media Journal: International Journal of Information and Communication Technology (IJICT) 2024 Vol.25 No.10 pp.62 - 76 Abstract: In the era of thriving new media, analysing student online sentiment is crucial for understanding student group dynamics and ensuring campus stability. This paper introduces a deep learning-based method for semantic understanding and categorisation of student online sentiment in new media. We collected extensive student speech data from social media, forums, and comments, creating a high-quality dataset through text pre-processing. A combined model that leverages both convolutional neural networks and long short-term memory networks efficiently captures textual characteristics and performs sentiment analysis. By incorporating an attention mechanism, the model focuses on key sentiment expressions. Experiments show our method's superiority in semantic understanding and categorisation tasks, with accuracy and F1 score improvements of approximately 15% and 18% over existing techniques, offering valuable insights for educational administrators and robust technical support for new media public opinion monitoring and analysis. Inderscience Publishers - linking academia, business and industry through research

Title: Deep learning semantic understanding and classification of student online public opinion for new media

Authors: Dan Wang; Li Wang

Addresses: Gingko College of Hospitality Management, Chengdu 611743, China ' Gingko College of Hospitality Management, Chengdu 611743, China

Abstract: In the era of thriving new media, analysing student online sentiment is crucial for understanding student group dynamics and ensuring campus stability. This paper introduces a deep learning-based method for semantic understanding and categorisation of student online sentiment in new media. We collected extensive student speech data from social media, forums, and comments, creating a high-quality dataset through text pre-processing. A combined model that leverages both convolutional neural networks and long short-term memory networks efficiently captures textual characteristics and performs sentiment analysis. By incorporating an attention mechanism, the model focuses on key sentiment expressions. Experiments show our method's superiority in semantic understanding and categorisation tasks, with accuracy and F1 score improvements of approximately 15% and 18% over existing techniques, offering valuable insights for educational administrators and robust technical support for new media public opinion monitoring and analysis.

Keywords: resource student online opinion; deep learning; semantic understanding; sentiment analysis; attention mechanism.

DOI: 10.1504/IJICT.2024.143335

International Journal of Information and Communication Technology, 2024 Vol.25 No.10, pp.62 - 76

Received: 16 Oct 2024
Accepted: 04 Nov 2024
Published online: 13 Dec 2024 *

Title: Deep learning semantic understanding and classification of student online public opinion for new media

Keep up-to-date