A Deep Hybrid Learning Approach to Detect Bangla Fake News

Adib, Quazi Adibur Rahman; Mehedi, Md Humaion Kabir; Sakib, Md. Sadman; Patwary, Kabbya Kantam; Hossain, Sabbir; Rasel, Annajiat Alim

doi:10.1109/ismsit52890.2021.9604712

Cited by 24 publications

(3 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The hybrid approach demonstrated superior performance compared to non-hybrid baseline approaches in the detection of fake news on two datasets, namely ISO and FA-KES. By inspiring this approach, authors [43] utilized a 1D CNN for the extraction of features and standard ML techniques for Bangla fake news classification. The proposed hybrid model obtained a 99.50% F1 score for overall data and 83.25% for fake labeled data for the random forest algorithm.…”

Section: Role Of Datasetsmentioning

confidence: 99%

Fake News Detection Datasets: A Review and Research Opportunities

Dhiman,

Kaur,

Hamid

et al. 2024

IJCDS

View full text Add to dashboard Cite

The impact of fake news is far-reaching, affecting journalism, the economy, and democracy. In response, there has been a surge in research focused on detecting and combating fake news, resulting in the development of datasets, techniques, and fact-verification methods. One crucial aspect of this effort is the creation of diverse and representative datasets for training and evaluating machine learning models for fake news detection. This review paper examines the available datasets relevant to detecting fake news, with a particular emphasis on those available in the Indian context, where few resources exist. By identifying research opportunities and highlighting existing corpora, this paper aims to assist researchers in improving their fake news detection studies and contributing to more comprehensive research on the topic. To the best of our knowledge, no survey has specifically focused on accessible corpora in the Indian context, making this review a valuable resource for researchers in the field.

show abstract

Section: Role Of Datasetsmentioning

confidence: 99%

Fake News Detection Datasets: A Review and Research Opportunities

Dhiman,

Kaur,

Hamid

et al. 2024

IJCDS

View full text Add to dashboard Cite

show abstract

“…In Late fusion, the features are extracted, and the fusion process happens [50]. The authors of [30] utilized the late fusion technique by utilizing Deep Neural Network (1D-CNN) as a feature extractor from the Bangla news and Machine Learning algorithms, namely, Random Forest, AdaBoost, K-Nearest Neighbours, SVM and Decision Tree for classification of fake and real news.…”

Section: Introductionmentioning

confidence: 99%

On the Effect of Intelligent Neurally Processed Features on COVID–19 Fake News Detection Performance

2024

nano-ntp

View full text Add to dashboard Cite

Online social networks are flooded with lot of user-generated information; fake news offenders use these online social network platforms to spread COVID fake news. This propagation of fake news results in a low level of content truthfulness, distrust in online social networks, panic and fear, which makes people take erroneous decisions. Hence, accurate classification of fake news against real news is mandatory. Therefore, in this study, a novel Neurally Augmented model is proposed to classify fake news accurately based on the content of the tweet. The proposed model creates a novel Deep Neural Network to automatically extract the neutrally processed features for downstream processing in a classic Machine Learning model. A neurally processed feature from 1D ConvNet is used to augment classic Machine Learning model in an attempt to improve the classification and discriminative capability of the classic Machine Learning model. Experimentation is done on a COVID-19 Twitter dataset curated by the authors. The proposed methodology provides a highly accurate fake news classification of 97.25%, which is 12% superior to the classic Machine Learning and Deep Learning models without neural augmentation. The proposed methodology is further evaluated on the ISOT, Indian Fake News Dataset, LIAR and Constraint Shared task COVID-19 Fake News benchmark datasets to determine its robustness, and it achieves significant accuracy of 93.75%, 89.17%, 83.68%, and 91%, respectively. The proposed model is tested on the proposed dataset and achieved a 5% increase in accuracy over the benchmark datasets.

show abstract

“…Though, the human-decision making can be contrived and sometimes influenced [17], blockchain technology in participation with smart-contract protocol has been proven effective for collaborative decision-making [18]. On the other hand, due to the lack of a proper representative dataset [19] and the morphological complexity [20], Bengali is a complex language to work on and achieve accuracy. In order to assess the veracity of the news, this study combines human reasoning capacity with a potent NLP language model.…”

Section: Introductionmentioning

confidence: 99%

Toward News Authenticity: Synthesizing Natural Language Processing and Human Expert Opinion to Evaluate News

et al. 2023

View full text Add to dashboard Cite

The growing popularity of online news has prompted concerns regarding (i) the socio-political influence over news dissemination, (ii) the waning freedom of news media, (iii) and a facile news evaluation process. A piece of news having the power to capture a large audience and sow the seed of bizarre consequences on a national scale should be prudently evaluated before reaching the mass. In quest of making a substantial profit, and sometimes due to inevitable socio-political influence, news with biased heading outpours mass media, resulting in ambiguity and mass manipulation. In this paper, we suggest a blockchain, smart contract, and incremental machine learning-based news evaluation procedure for the Bengali language to overcome these challenges. Weighted synthesis of machine classification and human expert opinion in a decentralized platform are synthesized to evaluate news. With continuous data, the Natural Language Processing (NLP) model is incrementally trained, and the best version of the model is used to detect deprived fake news. During experiments, the NLP model with initial training and testing accuracy of 84.94% and 84.99% was increased to 93.75% and 93.80% after nine rounds of incremental model training. On the Ethereum test network, the protocols have been installed and tested. The simulation demonstrates successful implementation of our proposed system.

show abstract

A Deep Hybrid Learning Approach to Detect Bangla Fake News

Cited by 24 publications

References 12 publications

Fake News Detection Datasets: A Review and Research Opportunities

Fake News Detection Datasets: A Review and Research Opportunities

On the Effect of Intelligent Neurally Processed Features on COVID–19 Fake News Detection Performance

Toward News Authenticity: Synthesizing Natural Language Processing and Human Expert Opinion to Evaluate News

Contact Info

Product

Resources

About