Year: 2024
Type: article
Abstract: In recent years, the amount of online shopping review data has increased dramatically. Obtaining information that helps business decision-making from such complex and massive reviews has become a difficult and important task for merchants. This paper uses sentiment analysis technology to innovatively introduce the attention mechanism on the LSTM infrastructure of the baseline model, and proposes a word vector structure and a BiGRU structure to build an online user sentiment analysis system based on deep learning. The system includes a user review sentiment classification and sentiment analysis model based on the improved GCN model. The experimental results also show the superiority of our method, which brings 4.73%, 7.84% and 5.72% F1-score improvements to the algorithm respectively. It proves that the two algorithms proposed in this paper can effectively achieve their goals and achieve high performance.
DOI: 10.4018/joeuc.355238
Authors: Ming Xu, Chen Peng, Yueyue Hou, Yating Xiao, Osama Sohaib
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4402667939