Unlock instant, AI-driven research and patent intelligence for your innovation.

Multi-field recommendation method based on multi-task learning

A technology of multi-task learning and recommendation method, applied in the field of cross-domain recommendation, which can solve the problems of low recommendation accuracy of single-domain recommendation system and user cold start.

Active Publication Date: 2021-05-11
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF8 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The present invention proposes a multi-domain recommendation method based on multi-task learning for the low recommendation accuracy of the single-domain recommendation system and the problem of user cold start

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Multi-field recommendation method based on multi-task learning
  • Multi-field recommendation method based on multi-task learning
  • Multi-field recommendation method based on multi-task learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0053] This embodiment proposes a multi-domain recommendation method based on multi-task learning. In the description of this embodiment, it should be noted that the present invention is applied to multiple domain recommendation methods, but for the convenience of description, two domains are used As a description of the implementation, named domain A and domain B, combined below figure 1 , figure 2 , image 3 be specified, such as figure 1 As shown, this embodiment specifically includes the following steps:

[0054] Step S1: Data preprocessing

[0055] Step S1.1: Obtain the scoring data of domain A and domain B; for user μ, if And the user μ has no scoring data in the domain t, then estimate the user’s rating data on the item according to the user’s behavior on the item; for the user μ, if head Then randomly select n items from the field t that the user μ does not belong to, and set the score to 0, that is, if a user only has scoring data in a certain field, but doe...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a multi-field recommendation method based on multi-task learning. For users in multiple fields, user dimension vectors in the multiple fields are extracted, implicit user vectors of the users in the multiple fields are synthesized through combination of a full-connection neural network and a cross-stitch multi-task learning network, real implicit interest vectors of the users are interactively mined, and the user recommendation accuracy in each field is improved. According to the method, a multi-task learning technology is used, interest preferences of users in all fields are fused, implicit interest vectors of the users in all the fields are enriched, and recommended articles can better meet user requirements; and if a certain single field lacks the user interaction data, the user cold start problem in the field can be solved according to user data in other fields.

Description

technical field [0001] The invention belongs to the field of cross-domain recommendation, and in particular relates to a multi-domain recommendation method based on multi-task learning. Background technique [0002] As an effective means of information overload, recommender systems have been widely used in various enterprise applications. With the development and growth of the enterprise, various applications will be launched to expand the product matrix of the enterprise, such as applications in the field of movies and applications in the field of books. In order to solve the user cold start problem of each system, the traditional approach is to migrate the application data with rich user interaction to the application field with sparse data using the method of knowledge transfer. There will be some problems in this way. If a user has a lot of interaction data in the data-sparse application field, but not much interaction data in the data-rich field, the traditional cross-...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06Q30/06G06F16/9535G06N3/08
CPCG06Q30/0631G06F16/9535G06N3/08Y02D10/00
Inventor 王杰江春华杨茂林吴济森陈宇涵谢凡
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA