False news detection method based on multi-modal fusion

A detection method and multi-modal technology, applied in character and pattern recognition, digital data information retrieval, still image data retrieval, etc., can solve the problems of missing important features, affecting accuracy, and weak generalization, and achieve training Fast speed, improved detection accuracy, and improved classification effect

Pending Publication Date: 2020-12-25
TIANJIN UNIV
View PDF1 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, these models only perform simple splicing and fusion of multimodal feature vectors in the detection module, and then directly use softmax as a simple classifier. In this way,...

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
  • False news detection method based on multi-modal fusion
  • False news detection method based on multi-modal fusion
  • False news detection method based on multi-modal fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0023] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

[0024] The invention provides a fake news detection method based on multimodal fusion. The whole method can be implemented in two stages, the first stage is the feature extraction of news data, and the other stage is the feature classification part. In the feature extraction of news data, the pre-trained models BERT and VGG19 are used to extract news data features, which can obtain deeper and more significant feature representations of news text data and news image data, and obtain fusion features by splicing the two. For the classification part, different from the previous structure that directly uses softmax as the classifier, the present invention improves the performance of d...

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 invention discloses a false news detection method based on multi-modal fusion. The false news detection method comprises the following steps of: (1) preprocessing news text data and news picture data of news; (2) extracting features of news data, specifically, using a news data feature extraction model composed of a pre-training model BERT model and a VGG19 model to perform feature extractionon the preprocessed news text and news picture data, and performing optimization training to obtain a trained news data feature extraction model; (3) performing data feature classification, specifically, inputting a training data set into the news data feature extraction model to obtain a news feature set, inputting the news feature set into a classifier as training data to perform training of a classification model, and completing model training based on the false news detection method.

Description

technical field [0001] The invention relates to the technical field of rumor detection, in particular to a multimodal fusion fake news detection method. Background technique [0002] With the increasing popularity of various social media, various types of information on social media, such as pictures, text or videos, have quickly become a hot spot for mass information consumption due to their fast dissemination, multiple acquisition channels, and low threshold for production. However, these characteristics also allow false news to spread widely. Due to the asymmetry of information, false news can be spread through social media to mislead vulnerable people who do not know the truth, can cause immeasurable negative effects, and even manipulate public opinion, false news It has become a major issue related to social stability. Therefore, rapid detection of false news is an important task. [0003] Earlier studies constructed classifiers by manually extracting text features, us...

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
IPC IPC(8): G06F16/33G06F16/35G06F16/55G06F16/583G06F40/30G06K9/62
CPCG06F16/3344G06F16/35G06F16/55G06F16/583G06F40/30G06F18/285G06F18/253
Inventor 刘爽潘云锋
Owner TIANJIN UNIV
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