Web service credible hybrid recommendation method considering timeliness
A hybrid recommendation and web service technology, applied in digital data information retrieval, special data processing applications, instruments, etc., can solve problems such as data sparsity and malicious recommendations
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0055] According to the time-sensitive Web service trusted hybrid recommendation method according to claim 1, it is characterized in that: the calculation of the time-sensitive user similarity includes the following steps:
[0056] (1) Let the set of users be U, U={u 1 ,u 2 ,...,u m}, the set of all services is S, S={s 1 ,s 2 ,...,s n}, with r i,k to represent user u i to service s k score, in order to eliminate the user's personal scoring habit, it is necessary to calculate the user u i to service s k The objective score of w i,k ,
[0057] (2) Considering the timeliness of user ratings, the closer the evaluation time is to the current time, the better it can reflect the user's preference, and the farther the evaluation time is from the current time, the smaller the reference value of the rating is; get the user u i to service s k The timeliness score of w time (i,k),
[0058]
[0059] Among them, w time (i, k) means user u i to service s k Timeliness of sc
Embodiment 2
[0065]According to the time-sensitive Web service trusted hybrid recommendation method according to claim 1, it is characterized in that: the calculation of the user similarity based on time-sensitive tags includes the following steps:
[0066] (1) First, establish a label set T={t according to the label information 1 ,t 2 ,...,t l}, user u i use tag t a The set of all marked services is S i,a ={s i1 ,s i2 ,...,s ib}; then calculate each user u according to the value of the user-item rating matrix i for each label t a The score value of , when the user uses the same label to label different services, the user's score for the label is the average of all service scores;
[0067] (2) When calculating the user's rating on the label, two factors that affect the label information must also be considered, that is, the timeliness of the label and the user's preference for the label, so the user u i pair label t a Rating V i,a The calculation formula is r i,k is user u i t
Embodiment 3
[0072] The time-sensitive trusted hybrid recommendation method for Web services is characterized in that: the calculation of user trust based on social networks includes the following steps:
[0073] (1) Calculate social relationship familiarity Fam(u i , u j ), I i,j means user u i and user u j The number of interactions, I imin means user u i Minimum number of interactions with other users, I imax means user u i the maximum number of interactions with other users;
[0074] (2) Calculate the core degree of social relationship Cor(u j ), Among them, Core(u j ) means user u j The core degree of social relationship, d j means user u j number of friends, d max Indicates the maximum number of friends of the user in the social relationship;
[0075] (3) Calculate the direct trust degree DT(u i , u j ), DT(u i , u j )=αFam((u i , u j )+βCore(u j ), α, β respectively represent the weight factors of familiarity and core degree in social relations, α+β=1;
[00
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.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap