Title: Precision marketing method for online information of new e-commerce products based on user tags

Authors: Min Li; Liyang Sun

Addresses: Shandong College of Electronic Technology, Shandong, Jinan, 250014, China ' Normal College of Yangzhou Vocational University, Jiangsu, Yangzhou, 225000, China

Abstract: To improve marketing effectiveness for new e-commerce products, a precise online information marketing method based on user tags is proposed. First, e-commerce user features are extracted through a hierarchical attention model. Second, user feature data are processed using the rotated forest and AdaBoost algorithm to establish user labels. Then, features of online information for new e-commerce products are extracted through information entropy and information gain calculations using the decision tree algorithm. Finally, precise marketing is achieved through semantic similarity between user tags and product features, along with implicit ratings from neighbouring users. Experiments show the Hamming distance of this method remains below 0.30, with adjusted Rand coefficient values ranging from 0.73 to 0.92, while the highest click-through rate for new products reaches 0.758.

Keywords: new e-commerce products; online information; precision marketing; user tags; feature extraction; semantic similarity.

DOI: 10.1504/IJCAT.2026.154039

International Journal of Computer Applications in Technology, 2026 Vol.78 No.4, pp.288 - 295

Received: 13 Feb 2025
Accepted: 10 Jun 2025
Published online: 10 Jun 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article