Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ IEEE Accessarrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
IEEE Access
Article . 2021 . Peer-reviewed
License: CC BY
Data sources: Crossref
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
IEEE Access
Article
License: CC BY NC ND
Data sources: UnpayWall
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
IEEE Access
Article . 2021
Data sources: DOAJ
https://dx.doi.org/10.60692/6a...
Other literature type . 2021
Data sources: Datacite
https://dx.doi.org/10.60692/rb...
Other literature type . 2021
Data sources: Datacite
DBLP
Article
Data sources: DBLP
versions View all 5 versions
addClaim

This Research product is the result of merged Research products in OpenAIRE.

You have already added 0 works in your ORCID record related to the merged Research product.

Rumour Detection Based on Graph Convolutional Neural Net

الكشف عن الشائعات بناءً على الشبكة العصبية التفافية للرسم البياني
Authors: Na Bai; Fei Meng; Xiaobin Rui; Zhixiao Wang;

Rumour Detection Based on Graph Convolutional Neural Net

Abstract

La détection de rumeurs est un sujet de recherche important dans les réseaux sociaux, et de nombreux modèles de détection de rumeurs sont proposés ces dernières années. Pour la tâche de détection de rumeurs, les informations structurelles d'une conversation peuvent être utilisées pour extraire des fonctionnalités efficaces. Cependant, de nombreux modèles de détection de rumeurs existants se concentrent sur les caractéristiques structurelles locales alors que les caractéristiques structurelles globales entre le tweet source et ses réponses ne sont pas utilisées efficacement. Pour tirer pleinement parti des caractéristiques structurelles globales et des informations de contenu, nous proposons un graphe de relation Source-Réponses (SR-graph) pour chaque conversation, dans lequel chaque nœud désigne un tweet, sa caractéristique de nœud est des vecteurs de mots pondérés et les arêtes désignent l'interaction entre les tweets. Sur la base de SR-graphes, nous proposons un Ensemble Graph Convolutional Neural Net avec un Mécanisme d'Allocation de Proportion de Nœuds (EGCN) pour la tâche de détection de rumeurs. Dans les expériences, nous vérifions d'abord que les caractéristiques structurelles extraites sont efficaces, puis nous montrons les effets de différentes dimensions d'incorporation de mots sur plusieurs indices de test. De plus, nous montrons que notre modèle EGCN proposé est comparable ou même meilleur que les modèles d'apprentissage automatique actuels.

La detección de rumores es un tema de investigación importante en las redes sociales, y en los últimos años se han propuesto muchos modelos de detección de rumores. Para la tarea de detección de rumores, la información estructural en una conversación se puede utilizar para extraer características efectivas. Sin embargo, muchos modelos de detección de rumores existentes se centran en las características estructurales locales, mientras que las características estructurales globales entre el tweet de origen y sus respuestas no se utilizan de manera efectiva. Para aprovechar al máximo las características estructurales globales y la información de contenido, proponemos Source-Replies relation Graph (SR-graph) para cada conversación, en la que cada nodo denota un tweet, su característica de nodo son vectores de palabras ponderados y los bordes denotan la interacción entre tweets. Con base en gráficos SR, proponemos una Red Neural Convolucional de Gráficos Conjuntos con un Mecanismo de Asignación de Proporción de Nodos (EGCN) para la tarea de detección de rumores. En los experimentos, primero verificamos que las características estructurales extraídas son efectivas, y luego mostramos los efectos de diferentes dimensiones de incrustación de palabras en múltiples índices de prueba. Además, mostramos que nuestro modelo EGCN propuesto es comparable o incluso mejor que los modelos actuales de aprendizaje automático de última generación.

Rumor detection is an important research topic in social networks, and lots of rumor detection models are proposed in recent years. For the rumor detection task, structural information in a conversation can be used to extract effective features. However, many existing rumor detection models focus on local structural features while the global structural features between the source tweet and its replies are not effectively used. To make full use of global structural features and content information, we propose Source-Replies relation Graph (SR-graph) for each conversation, in which every node denotes a tweet, its node feature is weighted word vectors, and edges denote the interaction between tweets. Based on SR-graphs, we propose an Ensemble Graph Convolutional Neural Net with a Nodes Proportion Allocation Mechanism (EGCN) for the rumor detection task. In experiments, we first verify that the extracted structural features are effective, and then we show the effects of different word-embedding dimensions on multiple test indices. Moreover, we show that our proposed EGCN model is comparable or even better than the current state-of-art machine learning models.

يعد اكتشاف الشائعات موضوعًا بحثيًا مهمًا في الشبكات الاجتماعية، وقد تم اقتراح الكثير من نماذج اكتشاف الشائعات في السنوات الأخيرة. بالنسبة لمهمة الكشف عن الشائعات، يمكن استخدام المعلومات الهيكلية في المحادثة لاستخراج ميزات فعالة. ومع ذلك، تركز العديد من نماذج الكشف عن الشائعات الحالية على الميزات الهيكلية المحلية بينما لا يتم استخدام الميزات الهيكلية العالمية بين تغريدة المصدر وردودها بشكل فعال. للاستفادة الكاملة من الميزات الهيكلية العالمية ومعلومات المحتوى، نقترح رسمًا بيانيًا لعلاقة المصدر بالردود (SR - graph) لكل محادثة، حيث تشير كل عقدة إلى تغريدة، وميزة العقدة الخاصة بها هي متجهات الكلمات المرجحة، والحواف تشير إلى التفاعل بين التغريدات. استنادًا إلى الرسوم البيانية للممثل الخاص، نقترح شبكة عصبية التفافية للرسم البياني للمجموعة مع آلية تخصيص نسبة العقد (EGCN) لمهمة الكشف عن الشائعات. في التجارب، نتحقق أولاً من أن السمات الهيكلية المستخرجة فعالة، ثم نعرض تأثيرات أبعاد تضمين الكلمات المختلفة على مؤشرات اختبار متعددة. علاوة على ذلك، نظهر أن نموذج EGCN المقترح لدينا قابل للمقارنة أو حتى أفضل من نماذج التعلم الآلي الحديثة الحالية.

Related Organizations
Keywords

Artificial intelligence, Public relations, Sociology and Political Science, FOS: Political science, Word (group theory), Word embedding, Social Sciences, Statistical Physics of Opinion Dynamics, Geometry, Structural engineering, Convolutional neural network, Node (physics), Rumor, Pattern recognition (psychology), Graph, Theoretical computer science, Engineering, Machine learning, Conversation, FOS: Mathematics, Political science, Consensus Formation, Rumour detection, word-vectors embedding, The Spread of Misinformation Online, Statistical and Nonlinear Physics, Linguistics, Rumor Detection, graph convolutional neural nets, Computer science, TK1-9971, FOS: Philosophy, ethics and religion, Philosophy, Physics and Astronomy, Physical Sciences, FOS: Languages and literature, Statistical Mechanics of Complex Networks, Electrical engineering. Electronics. Nuclear engineering, Mathematics, Embedding

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    24
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Top 10%
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Top 10%
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Top 10%
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
24
Top 10%
Top 10%
Top 10%
gold