Research of Recommendation System Based on Deep Interest Network

With the development of e-commerce, it provides users with many choices. But how to quickly and accurately recommend goods to users is an important topic in this field. Matrix factorization recommendation model based on scoring data is widely used, but data sparsity affects the recommendation qualit...

Full description

Saved in:
Bibliographic Details
Published in:Journal of physics. Conference series 2021-01, Vol.1732 (1), p.12015
Main Authors: Wang, Shan-Shan, Pan, Yuan-Yuan, Yang, Xiao
Format: Article
Language:English
Subjects:
Algorithms
Physics
Recommender systems
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With the development of e-commerce, it provides users with many choices. But how to quickly and accurately recommend goods to users is an important topic in this field. Matrix factorization recommendation model based on scoring data is widely used, but data sparsity affects the recommendation quality of the model. In the paper, a product recommendation algorithm based on deep interest network is proposed. First, the user's purchased goods and the user's search are embedded coded, and the sparse features are transformed into low-dimensional dense features.Then, the feature vectors of the purchased goods and the search text vectors are joined together to input the deep interest network as the feature to predict the user's interest.Finally, the effectiveness of the model is validated by using record data. The validity of the model is verified by comparing with other models.
ISSN:1742-6588
1742-6596
DOI:10.1088/1742-6596/1732/1/012015