E-commerce recommender system design based on web information extraction and sentiment analysis

Jinfeng Feng

PLOS ONE, 2025, vol. 20, issue 9, 1-19

Abstract: The research proposes an e-commerce recommendation system based on web page information extraction and sentiment analysis. Through the improved S-PageRank algorithm and the dynamic topic library generation strategy, the precision rate of cross-platform commodity information extraction has been significantly improved to 90%, which is superior to the traditional S-PageRank algorithm. The template-based web page information extraction method performs well, with a precision rate 10% higher than that of the method based on the document object model. In terms of sentiment analysis, the comprehensive attention mechanism model combining the topic model and the bidirectional long short-term memory network has achieved the precise calculation of the sentiment scores of each topic in the customer evaluation. When the number of topics of the LDA model is 7, the prediction accuracy reaches its peak, and the model outperforms previous methods in terms of accuracy, recall rate and F-score. The experimental results show that this recommendation system performs excellently in the prediction of sentiment trends and the analysis of the reasons behind emotions. Its prediction accuracy and analysis accuracy are both superior to existing recommendation systems such as Amazon and Netflix. This system can provide users with more accurate and personalized product recommendation services, and at the same time offer merchants deeper insights into users’ emotions.

Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0327213 (text/html)
https://journals.plos.org/plosone/article/file?id= ... 27213&type=printable (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:plo:pone00:0327213

DOI: 10.1371/journal.pone.0327213

Access Statistics for this article

More articles in PLOS ONE from Public Library of Science
Bibliographic data for series maintained by plosone ().