Social media content classification and community detection using deep learning and graph analytics

Ali, Mohsan; Hassan, Mehdi; Kifayat, Kashif; Kim, Jin Young; Hakak, Saqib; Khan, Muhammad Khurram

doi:10.1016/j.techfore.2022.122252

Cited by 17 publications

(10 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ali. [10] provides fresh deep learning as well as graph-based algorithms for detecting hate speech, followed by approaches for detecting communities and researching social media for hate content detection. Hate material is classi ed by using the suggested modi ed LSTMGRU model.…”

Section: Related Workmentioning

confidence: 99%

Enhancing Communication and Comprehension for Individuals with Special Needs through Federated Learning: A Deep Learning Approach

Elsayed,

Elrashidy,

EL-Sayed

et al. 2023

Preprint

View full text Add to dashboard Cite

Individuals with special needs most of the time find it harder to identify hazards and dangers as well as circumstances that are socially challenging. Hence, they face the risk of falling victim to abuse and violence. In this paper, the main goal is to help people with special needs to more successfully communicate with others and comprehend their surroundings. Machine learning-based solutions are used to help people with special needs in their communication tasks. The proposed machine learning model contains a convolutional layer, attention layer, and Bidirectional long short-term memory (BiLSTM) layer and achieves 99.00% accuracy performance. We applied federated learning to preserve privacy and to help researchers overcome problems they face when dealing with people with special needs.

show abstract

Section: Related Workmentioning

confidence: 99%

Enhancing Communication and Comprehension for Individuals with Special Needs through Federated Learning: A Deep Learning Approach

Elsayed,

Elrashidy,

EL-Sayed

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Studies have shown that aggressive behavior is prevalent on Twitter (Kwak et al, 2010;Wu et al, 2011). Several studies have created annotated datasets of aggression in various languages, such as English (Kumar et al, 2018b,a;Bhattacharya et al, 2020;Kumar et al, 2020;Ali et al, 2023), Hindi (Kumar et al, 2018b,a;Bhattacharya et al, 2020;Kumar et al, 2020), Bengali (Bhattacharya et al, 2020;Kumar et al, 2020;Sharif and Hoque, 2022), Italian (Gattulli et al, 2022), Spanish (Torregrosa et al, 2022), Turkish (Balci and Salah, 2015), and Russian (Gordeev, 2016). These datasets were curated from social media platforms such as Facebook, Twitter, YouTube, and others.…”

Section: Aggression Detectionmentioning

confidence: 99%

“…Di erent deep-learning models for aggression detection have also been used. These include CNN (Agbaje and Afolabi, 2022), LSTM (Agbaje and Afolabi, 2022;Aroyehun and Gelbukh, 2018;Kumari et al, 2021;Pareek et al, 2022;Ali et al, 2023), BiLSTM (Srivastava and Khurana, 2019), and their combinations. Di erent features such as two-dimensional TF-IDF vectors (Chen et al, 2020), embedding from Convolutional Capsule Layer (Srivastava and Khurana, 2019) and embedding from FastText (Pareek et al, 2022), and sentiment analysis (Agbaje and Afolabi, 2022) have also been employed for aggression detection.…”

Section: Aggression Detectionmentioning

confidence: 99%

“…Multilayer perceptron classifiers with TF-IDF of unigram and bigram features are e ective in identifying aggression (Sadiq et al, 2021). Moreover, LSTM with GRU (Ali et al, 2023), and deep learning models with emotional features and word embeddings have shown better performance (Khan et al, 2022). An ensemble of multiple fine-tuned BERT models based on bootstrap aggregating for aggression detection was proposed by Risch and Krestel (2020), which performed best on the dataset developed by Kumar et al (2020).…”

Section: Aggression Detectionmentioning

confidence: 99%

“…The General Aggression Model (GAM) developed by Allen et al (2018) explores the factors that influence aggressive behavior. (Ali et al, 2023) has used the Girvan Newman community detection algorithm to detect aggressive communities of social media influencers. Henneberger et al (2017) found that users can be influenced to act aggressively and bully others because of high toxicity and aggression in their social circle.…”

Section: Aggression Behaviormentioning

confidence: 99%

See 2 more Smart Citations

You are what your feeds makes you: A study of user aggressive behaviour on Twitter

Mane¹,

Kundu²,

Sharma³

2023

Preprint

View full text Add to dashboard Cite

<p>Social media has become an essential part of human life, providing people with a means of communication and information dissemination. However, the rise of antisocial behavior on these platforms can create a sense of anonymity within various groups, which may have negative consequences for society as a whole. At the individual level, aggressive behavior on social networks has become a concern, as it can lead to negative consequences, such as mental harm. This study aims to understand whether exposure to aggressive content can make users more aggressive. To investigate this, we analyzed a large dataset of Twitter posts, consisting of 15 million tweets from 1.5 million users. We manually annotated 6,000 English posts and designed a neural network based aggression detection model with 81\% accuracy. We then measure users’ aggressive behavior using a novel metric, ``user aggression intensity''. This metric takes into account the overall aggressive activity of users. By utilizing this metric, we profile users as aggressive or non-aggressive. The analysis of aggressive tweets of users and their feeds suggests that the content of one's social media feeds can influence aggressive behavior, and event-specific feeds can lead to increased aggression. Furthermore, the study revealed that users tend to support and encourage aggressive content on social media, which can contribute to the proliferation of antisocial behavior. Understanding the underlying causes of such behavior can help design effective strategies to mitigate its negative effects and promote a healthy social media environment.</p>

show abstract

Topological to deep learning era for identifying influencers in online social networks :a systematic review

Rashid

Bhat

2023

Multimed Tools Appl

View full text Add to dashboard Cite

Social media content classification and community detection using deep learning and graph analytics

Cited by 17 publications

References 34 publications

Enhancing Communication and Comprehension for Individuals with Special Needs through Federated Learning: A Deep Learning Approach

Enhancing Communication and Comprehension for Individuals with Special Needs through Federated Learning: A Deep Learning Approach

You are what your feeds makes you: A study of user aggressive behaviour on Twitter

Topological to deep learning era for identifying influencers in online social networks :a systematic review

Contact Info

Product

Resources

About