A social media rumor detection method based on multi-task learning

A multi-task learning and social media technology, applied in the field of social media rumor detection based on multi-task learning, can solve problems such as difficult detection, achieve the effect of increasing training samples, reducing over-fitting phenomenon, and enhancing performance

Active Publication Date: 2021-05-25
长沙市智为信息技术有限公司
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The task of rumor detection on social media is challenging since traditional news media detection algorithms are ineffective or inapplicable to the task of rumor detection in social media, and it is difficult to detect cases where rumors are intentionally written to mislead readers.

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 social media rumor detection method based on multi-task learning
  • A social media rumor detection method based on multi-task learning
  • A social media rumor detection method based on multi-task learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0049] A social media rumor detection method based on multi-task learning, specifically comprising the following steps:

[0050] S1: Perform data extraction and format conversion on the corpus in the social media text dataset, and obtain the source, reply and propagation path of the post;

[0051] S2: Extract the features of the writing style from the corpus processed in step 1, and process it in the form of a vector;

[0052] S3: Extract the feature of user confidence from the corpus processed in step 1, and process it in the form of a vector;

[0053] S4: Do text preprocessing on the text part of the source post and the reply post, and encode the text into a vector form as a text representation to input into the follow-up task;

[0054] S5: Concatenate the features extracted by S2 and S3 with the text representation of S4;

[0055] S6: Put the spliced ​​vectors into a shared BERT layer, and encode the data of subtask I position detection and subtask II rumor detection into...

Embodiment 2

[0086] (1) For the rumor detection of social media, this invention proposes a model method based on multi-task joint learning, which is used to automatically detect the authenticity of post content in social media and avoid the "post-truth" problem caused by rumors.

[0087] (2) The present invention divides the task of rumor detection in social media into two tasks: classifying the standpoints (support, objection, question, statement) of the participants on the post and classifying the authenticity (true, false, neutral) of the post statement itself. subtasks.

[0088](3) Since the accuracy of the posts is strongly correlated with the attitudes of the participants to the posts, the model establishes two tasks to learn together, share parameters, and inspire each other, so that the features learned by the two tasks are more generalizable , and finally evaluate the authenticity of the post.

[0089] (4) The present invention adds features in the preprocessing part, including t...

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 social media rumor detection method based on multi-task learning, which specifically includes the following steps: S1: performing data extraction and format conversion on the corpus, obtaining the source, reply and propagation path of the post; S2: extracting the characteristics of the writing style; S3: Extract the features of user confidence; S4: Do text preprocessing on the text part of the source post and the reply post to input the follow-up tasks; S5: Concatenate the features extracted by S2 and S3 with the text representation of S4; S6: Put the spliced ​​vectors into a shared BERT layer; S7: Construct the neural network structure separately; S8: Input the data processed by S5 into the neural network structure, and output position classification and rumor classification. The invention can combine two highly related tasks with a multi-task joint model, improves the tasks of rumor detection and position classification, and improves the performance of rumor detection.

Description

technical field [0001] The invention relates to the technical field of rumor detection, in particular to a multi-task learning-based social media rumor detection method. Background technique [0002] In recent years, with the rapid development of social media, people tend to check relevant news they care about through social media such as twitter and reddit. However, while these social media provide convenience to our life, they also lead to the problem of flooding of information and spreading rumors on the Internet. Rumors have brought a lot of harm to people's production and life. Viral rumors often arouse public opinion, disrupt social order, and bring negative impacts on social economy and politics. At the same time, rumors can also affect people's judgment. [0003] The bad influence of rumors has aroused widespread public concern, and rumor detection technology needs to be improved urgently. The task of rumor detection on social media is challenging since traditiona...

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): G06F16/33G06F16/35G06F16/9536G06F40/205G06F40/216G06F40/284G06F40/30G06K9/62G06N3/04G06N3/08G06Q50/00
Inventor 李芳芳张盼曦宁肯刘志
Owner 长沙市智为信息技术有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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