Account recommendation method and device
A recommendation method and account technology, applied in the computer field, can solve the problems of limited scope and single account recommendation method, and achieve the effect of expanding the scope and improving the diversity.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment approach
[0057] It can be understood that the recommended account may be at least one target account in the account candidate set. The number of specific recommended accounts may be determined according to actual needs. In one embodiment, the recommended accounts may be all target accounts included in the account candidate set.
[0058] In the embodiment of this specification, by recalling target accounts satisfying at least one recall rule according to the multi-channel recall rules, and then determining the account candidate set according to the recalled target accounts, and sending at least one target account included in the account candidate set to the target user, Compared with the single recommendation method in the prior art that only recommends based on user interests, since the target accounts in the account candidate set are recalled based on multi-channel recall rules, it is possible to recommend accounts to users from multiple angles and improve the diversification of accoun...
specific Embodiment approach 1
[0061] In one or more embodiments of this specification, when the multi-channel recall rule is a user coordinated recall rule, recalling the target account according to the multi-channel recall rule may include:
[0062] According to the collaborative filtering algorithm, users with similar interests to the target user are determined;
[0063] Recall target accounts corresponding to users with similar interests to the target user.
[0064] In practical applications, user collaboration rules may include determining users having similar interests to the target user according to interest similarity, specifically, determining users having similar interests to the target user according to a collaborative filtering algorithm. The collaborative filtering algorithm can mainly use the similarity of user behavior to calculate the similarity of interest. For example, given user u and user v, N(u) can be expressed as a collection of items that user u has had positive feedback, and N(v) ca...
specific Embodiment approach 2
[0070] In practical applications, considering that the target user may be an advertiser user, and the historical behavior of the advertiser user may include placing an advertisement on a specified account, then in one or more embodiments of this specification, the multi-channel recall rule is the user historical behavior When recalling the rules, then, according to the multi-channel recall rules, recalling the target account can include:
[0071] Recalling target accounts that have been served advertisements by target users within the first preset historical time period.
[0072] As mentioned above, considering that the target user may be an advertiser user, the historical behavior of the target user may include that the target user has placed an advertisement on a specified account, and the specified account may be a target account that has been placed an advertisement by the target user.
[0073] It is understandable that, considering the timeliness, it is not necessary to r...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com