Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A topology-free method for predicting the propagation range of social messages

A prediction method and message technology, which is applied in the field of social message propagation range prediction, can solve the problem of not considering the mutual influence of messages, and achieve the effect of improving accuracy

Active Publication Date: 2022-03-04
上海帮赋成科技有限公司
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0019] The purpose of the present invention is to solve the problem that the existing methods do not consider the mutual influence of messages during the propagation process, and propose a topology-free social message propagation range prediction method

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 topology-free method for predicting the propagation range of social messages
  • A topology-free method for predicting the propagation range of social messages
  • A topology-free method for predicting the propagation range of social messages

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0047] Embodiment 1: The specific process of a method for predicting the dissemination range of social messages without a topology structure in this embodiment is as follows:

[0048] The invention studies the problem of message propagation range prediction under the condition of no topology structure, and proposes a topology-free message propagation range prediction method NT-EP. The method consists of 4 parts: (1) Construct a weighted propagation graph for each message by using the characteristics of message propagation decay with time, use random walk strategy to obtain multiple propagation paths on the propagation graph, and then use the word2vec method to calculate each message. (2) Replace the propagation path of the target message with the user's feature vector sequence and input it to the bidirectional gated recurrent neural network (Bi-GRU), and combine the attention mechanism to calculate the propagation feature vector of the target message; ( 3) Considering the poss...

specific Embodiment approach 2

[0060] Embodiment 2: The difference between this embodiment and Embodiment 1 is that in step 1, a weighted propagation graph is constructed for each message according to the propagation time difference in the message action log, as shown in the following example: figure 1 shown in (a) and (b). The numbers on the side of the propagation graph represent the probability of influence between users. After the propagation graph is constructed, a random walk method is used to extract several possible propagation paths of the message from the propagation graph, such as figure 1 As shown in (c); the specific process is:

[0061] Propagation path selection

[0062] A given action log usually sorts the actions of each message by propagation time, as in figure 2 shown in (a). User V 1 Message A was received at time 1 1 , user V 2 Message A was accepted at time 2 1 , . . . The true propagation trajectory of the message cannot be obtained from the given action log. Because the rea...

specific Embodiment approach 3

[0071] Embodiment 3: The difference between this embodiment and Embodiment 1 or 2 is that in the second step, the word2vec method is used to calculate the initial feature vector of each target user on the propagation path of the message; the specific process is:

[0072] After extracting the propagation paths of all messages, each propagation path is regarded as a sentence, and each user on the path is regarded as a word in the sentence, and input to word2vec [23] (Le Q, Mikolov T. Distributed representations of sentences and documents[C] / / International conference on machine learning. 2014:1188-1196.) In the skip-gram model, the initial feature vector of each target user is obtained; assuming that the user's initial The dimension of the feature vector is H.

[0073] Other steps and parameters are the same as in the first or second embodiment.

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

A method for predicting the propagation range of social information without topology, and the invention relates to a method for predicting the propagation range of social information. The purpose of the present invention is to solve the problem that the existing methods do not consider the mutual influence of messages during the propagation process. The process is: 1. After the propagation graph is constructed, use random walk to extract several possible propagation paths of the message from the propagation graph; 2. Use the word2vec method to calculate the initial characteristics of each target user on the propagation path of the message Vector; 3. Get the final vector representation of each target user on the propagation path; 4. Calculate the propagation feature vector of each target message; 5. Calculate the influence vector of other messages; 6. Spread the target message obtained in 4. The eigenvectors are combined with the influence vectors of other messages to fit the incremental propagation range of the target message using MLP. The invention is used in the field of message propagation range prediction.

Description

technical field [0001] The present invention relates to a social message dissemination range prediction method. Background technique [0002] With the rapid development of social networks in recent years, more and more users use Sina Weibo, Twitter, Facebook and other social networking sites to share their life. According to statistics, Facebook has more than 2.3 billion monthly active users as of December 31, 2018 [1] (Zephoria. The top 20valuable Facebook statistics-up-dated April 2018. [Online], Available: https: / / zephoria.com / top-15-valuable-facebook-statistics / , January 1, 2019.). This shows that social networking has become a part of many people's lives. At the same time, major social platforms are also promoting the rapid dissemination of various news. For example, hundreds of millions of microblogs are generated on Sina Weibo every day. There will be a lot of important information in Weibo generated every day. A user's update of a Weibo may contain the user's at...

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): H04L41/147H04L51/52
CPCH04L41/147G06Q10/04G06Q50/01H04L51/52
Inventor 刘勇刘子图李晓坤
Owner 上海帮赋成科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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