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

A service recommendation method and system based on transfer learning

A technology for service recommendation and transfer learning, applied in the field of service recommendation methods and systems based on transfer learning, to avoid overfitting, improve training speed, and improve security.

Active Publication Date: 2022-03-04
SHANDONG UNIV
View PDF12 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] Due to the exposure of privacy issues in recent years and the gradual increase of users' security awareness, people are more inclined to transmit as little personal user data as possible to the server, so that the recommendation system can only obtain relatively sparse user data, but The system still needs to recommend suitable services to users in this environment, which is the cold start problem

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
  • A service recommendation method and system based on transfer learning
  • A service recommendation method and system based on transfer learning
  • A service recommendation method and system based on transfer learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0070] A service recommendation method based on transfer learning, such as figure 1 shown, including the following steps:

[0071] S1. Establish a data set: collect relevant data on the Internet, filter the valid data in the data, and establish a data set;

[0072] The relevant data collected online is divided into two parts, one part is directly obtained from http: / / www2.informatik.uni-freiburg.de / ~cziegler / BX / The data set obtained by downloading contains user information, item information, and user rating information for different items. The other part is based on the item number sequence in the above data, from https: / / www.amazon.com The corresponding descriptive information for different items is captured as supplementary information data.

[0073] Specifically, during the crawling process, a python script is used, based on the BeautifulSoup library and the selenium library, to simulate the behavior of the browser, and at the same time to access the Amazon website...

Embodiment 2

[0081] According to a service recommendation method based on transfer learning described in Embodiment 1, the difference is that:

[0082] In step S2, service semantic modeling, the specific implementation steps are as follows:

[0083] S201. Perform dataset labeling work:

[0084] Analyzing Textual Content Information in Datasets Using Hidden Dirichlet Distribution Info c and geographic location information Info l ; The text content information refers to the descriptive information captured from the Amazon website in the data set established in step S1. The geographic location information refers to the geographic location when different users request different items in the dataset directly downloaded from the dataset created in step S1.

[0085] The unsupervised clustering algorithm using the hidden Dirichlet distribution is used to calculate the similar topic model of each service, and the description information is projected and mapped to a vector space composed of multi...

Embodiment 3

[0130] A service recommendation system based on migration learning, used to implement the service recommendation method based on migration learning described in Embodiment 1 or 2, including a data set establishment module, a service semantic modeling module, a migration learning module, a matrix decomposition module, a service Recommended modules;

[0131] The data set building module is used to execute step S1; collect data such as user and service description information; the service semantic modeling module is used to execute step S2; the data used for preliminary processing, including the introduction of geographic location information, for semantic modeling , to obtain different themes; the migration learning module is used to execute step S3; the obtained preliminary semantic processing results are combined with the information of different neighborhoods for migration to obtain more detailed description results; the matrix decomposition module is used to execute step S4; ...

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 present invention relates to a service recommendation method and system based on migration learning, including: S1. Establishing a data set: collecting relevant data, and filtering valid data to establish a data set; S2. Service semantic modeling: extracting different information from the data set The geographic location information in the database is clustered and analyzed, and the appropriate ratio is used as a feature to fuse it into the original information to perform service semantic modeling; S3, transfer learning: combine cross-domain prior information and service semantic modeling information Refine it; S4. Decomposition of data fusion matrix: After obtaining the correction coefficient, determine the hidden factors of the user-theme matrix and the theme-service matrix, and perform different weight decay techniques on the a priori determined hidden factors, using the ADAM optimizer To combat over-fitting; S5, perform service recommendation for users. The present invention is used to provide stable service recommendation quality in a complex environment, and can also provide users with good service experience in a hot start environment.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, in particular to a transfer learning-based service recommendation method and system. Background technique [0002] Due to the exposure of privacy issues in recent years and the gradual increase of users' security awareness, people are more inclined to transmit as little personal user data as possible to the server, so that the recommendation system can only obtain relatively sparse user data, but The system still needs to recommend suitable services to users in this environment, which is the cold start problem. [0003] In a cold start environment, only a small amount of user data (such as GPS, biometric fingerprints, phone calls, historical data, etc.) can be obtained. Therefore, how to complete service recommendations for users based on existing sparse information and provide stable and Effective quality of service has become an urgent problem to be solved. Contents o...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/335G06F16/33G06F16/35G06F16/387G06F40/289G06F40/30G06K9/62G06N20/00
CPCG06F16/335G06F16/35G06F16/3346G06F16/387G06F40/289G06F40/30G06N20/00G06F18/23213
Inventor 戴鸿君雷超
Owner SHANDONG UNIV