Title:
CONTENT RECOMMENDATION METHOD BASED ON HETEROGENEOUS FEATURE DEEP RESIDUAL NETWORK
Document Type and Number:
WIPO Patent Application WO/2022/088417
Kind Code:
A1
Abstract:
A content recommendation method based on a heterogeneous feature deep residual network. The method comprises: obtaining source data of a content to be recommended, and performing data conversion processing on the source data to obtain weighted hybrid embedded data (S100); obtaining predictive scoring data according to the weighted hybrid embedded data and a deep residual network model (S200); and obtaining a recommendation result of the content according to the predictive scoring data (S300). The source data of a domain content is processed, and the processed data is inputted into the residual network model to obtain the predictive scoring data, and then the domain content is accurately recommended according to the predictive scoring data, such that the calculation method is high in efficiency, and low in resource occupancy. In addition, the use of diversified data can also avoid the occurrence of cold start when recommendation is performed for new users.
Inventors:
CAI SHUBIN (CN)
MING ZHONG (CN)
ZHOU HUAIFENG (CN)
PENG TAO (CN)
MING ZHONG (CN)
ZHOU HUAIFENG (CN)
PENG TAO (CN)
Application Number:
PCT/CN2020/136144
Publication Date:
May 05, 2022
Filing Date:
December 14, 2020
Export Citation:
Assignee:
UNIV SHENZHEN (CN)
International Classes:
G06F16/9535
Foreign References:
| CN109726806A | 2019-05-07 | |||
| CN109933678A | 2019-06-25 | |||
| CN110674265A | 2020-01-10 | |||
| CN111651613A | 2020-09-11 | |||
| US10360900B1 | 2019-07-23 |
Attorney, Agent or Firm:
JOHNSON INTELLECTUAL PROPERTY AGENCY (SHENZHEN) (CN)
Download PDF: