Deep vs. Shallow: A Comparative Study of Machine Learning and Deep Learning Approaches for Fake Health News Detection DOI Creative Commons
Tripti Mahara, Helen Josephine V L,

Rashmi Srinivasan

и другие.

IEEE Access, Год журнала: 2023, Номер 11, С. 79330 - 79340

Опубликована: Янв. 1, 2023

The internet explosion and penetration have amplified the Fake news problem that existed even before penetration. This becomes more of a concern, if it is health-related news. To address this issue, research aims to propose Content based (CBM) Feature Based Models (FBM). difference between both models in input provided. CBM only takes content as whereas FBM along with contents also two readability features input. Under each category performance five traditional machine learning techniques: Decision Tree, Random Forest, Support Vector Machine, Adaboost-Decision Tree Adaboost-Random Forest compared hybrid Deep Learning approaches namely CNN-LSTM CNN-BiLSTM. News Healthcare data set comprising 9581 articles utilized for study. As highly imbalanced dataset, Easy Data Augmentation technique used balance dataset. Experimental results demonstrate performed better than Models. Amongst proposed FBM, Hybrid CNN - LSTM model had F1 score 97.09% Score 98.9%. Thus under best performing classification fake

Язык: Английский

Detection of Fake News Text Classification on COVID-19 Using Deep Learning Approaches DOI Open Access

Waqas Haider Bangyal,

Rukhma Qasim,

Najeeb Ur Rehman

и другие.

Computational and Mathematical Methods in Medicine, Год журнала: 2021, Номер 2021, С. 1 - 14

Опубликована: Ноя. 15, 2021

A vast amount of data is generated every second for microblogs, content sharing via social media sites, and networking. Twitter an essential popular microblog where people voice their opinions about daily issues. Recently, analyzing these the primary concern Sentiment analysis or opinion mining. Efficiently capturing, gathering, sentiments have been challenging researchers. To deal with challenges, in this research work, we propose a highly accurate approach SA fake news on COVID-19. The dataset contains COVID-19; started by preprocessing (replace missing value, noise removal, tokenization, stemming). We applied semantic model term frequency inverse document weighting representation. In measuring evaluation step, eight machine-learning algorithms such as Naive Bayesian, Adaboost, K -nearest neighbors, random forest, logistic regression, decision tree, neural networks, support vector machine four deep learning CNN, LSTM, RNN, GRU. Afterward, based results, boiled efficient prediction python, trained evaluated classification according to performance measures (confusion matrix, rate, true positives rate...), then tested set unclassified COVID-19, predict sentiment class each Obtained results demonstrate high accuracy compared other models. Finally, recommendations provided future directions help researchers select data.

Язык: Английский

Процитировано

104
Detection and moderation of detrimental content on social media platforms: current status and future directions DOI Open Access
Vaishali U. Gongane, Mousami V. Munot, Alwin Anuse

и другие.

Social Network Analysis and Mining, Год журнала: 2022, Номер 12(1)

Опубликована: Сен. 5, 2022

Язык: Английский

Процитировано

66
Applications of Artificial Intelligence, Machine Learning, Big Data and the Internet of Things to the COVID-19 Pandemic: A Scientometric Review Using Text Mining DOI Open Access
Ignacio Rodríguez‐Rodríguez, José‐Víctor Rodríguez, Niloofar Shirvanizadeh

и другие.

International Journal of Environmental Research and Public Health, Год журнала: 2021, Номер 18(16), С. 8578 - 8578

Опубликована: Авг. 13, 2021

The COVID-19 pandemic has wreaked havoc in every country the world, with serious health-related, economic, and social consequences. Since its outbreak March 2020, many researchers from different fields have joined forces to provide a wide range of solutions, support for this work artificial intelligence (AI) other emerging concepts linked intelligent data analysis been decisive. enormous amount research high number publications during period makes it difficult obtain an overall view applications AI management understanding how field evolving. Therefore, paper, we carry out scientometric area supported by text mining, including review 18,955 related Scopus database 2020 June 2021 inclusive. For purpose, used VOSviewer software, which was developed at Leiden University Netherlands. This allowed us examine exponential growth on issue distribution country, highlight clear hegemony United States (USA) China respect. We automatic process extract topics interest observed that most important current lines focused patient-based solutions. also identified relevant journals terms pandemic, demonstrated growing value open-access publication, highlighted influential authors means citations co-citations. study provides overview status application pandemic.

Язык: Английский

Процитировано

58
A review on fake news detection 3T’s: typology, time of detection, taxonomies DOI Open Access
Shubhangi Rastogi, Divya Bansal

International Journal of Information Security, Год журнала: 2022, Номер 22(1), С. 177 - 212

Опубликована: Ноя. 15, 2022

Язык: Английский

Процитировано

47
OptNet-Fake: Fake News Detection in Socio-Cyber Platforms Using Grasshopper Optimization and Deep Neural Network DOI
Sanjay Kumar, Akshi Kumar, Abhishek Mallik

и другие.

IEEE Transactions on Computational Social Systems, Год журнала: 2023, Номер 11(4), С. 4965 - 4974

Опубликована: Март 3, 2023

Exposure to half-truths or lies has the potential undermine democracies, polarize public opinion, and promote violent extremism. Identifying veracity of fake news is a challenging task in distributed disparate cyber-socio platforms. To enhance trustworthiness on these platforms, this article, we put forward detection model, OptNet-Fake. The proposed model architecturally hybrid that uses meta-heuristic algorithm select features based usefulness trains deep neural network detect social media. $d$ -D feature vectors for textual data are initially extracted using term frequency inverse document (TF-IDF) weighting technique. then directed modified grasshopper optimization (MGO) algorithm, which selects most salient text. selected fed various convolutional networks (CNNs) with different filter sizes process them obtain notation="LaTeX">$n$ -gram from These finally concatenated news. results evaluated four real-world datasets standard evaluation metrics. A comparison algorithms recent methods also done. distinctly endorse superior performance OptNet-Fake over contemporary models across datasets.

Язык: Английский

Процитировано

27
Fake news detection: Taxonomy and comparative study DOI
Faramarz Farhangian, Rafael M. O. Cruz, George D. C. Cavalcanti

и другие.

Information Fusion, Год журнала: 2023, Номер 103, С. 102140 - 102140

Опубликована: Ноя. 14, 2023

Язык: Английский

Процитировано

27
Fake News in Virtual Community, Virtual Society, and Metaverse: A Survey DOI
Jinxia Wang,

Stanislav Makowski,

Alan Cieślik

и другие.

IEEE Transactions on Computational Social Systems, Год журнала: 2023, Номер 11(4), С. 4828 - 4842

Опубликована: Фев. 6, 2023

In the trend of accelerated progression communication network technology, emergence virtual communities (VCs), societies (VSs), metaverse, and other technologies not only makes data access sharing easier but also leads to proliferation fake news (FN). To effectively monitor identify FN in VC, VS, create a safer space, this work takes metaverse as objects. First, content display methods are reviewed explained, it is understood that mainly displayed by single-modal multimodal representations. Second, application scenarios many important fields such transportation analyzed, so further understand impact detection effect different scenarios. Finally, an intelligent outlook summary analysis carried out on information security FN, which provides theoretical reference new opportunities for identification cyberspace.

Язык: Английский

Процитировано

26
Combating the menace: A survey on characterization and detection of fake news from a data science perspective DOI Creative Commons
Wazib Ansar,

Saptarsi Goswami

International Journal of Information Management Data Insights, Год журнала: 2021, Номер 1(2), С. 100052 - 100052

Опубликована: Ноя. 1, 2021

Journalism has always remained a vital constituent of our society and journalists play key role in making people aware the happenings developments society. This spread information enables shaping ideologies, orientations thoughts individuals as well Contrary to this, misinformation or fake news leads detrimental consequences. With advent social media, menace become grievous due unrestrained propagation difficulty track several accounts operated by humans bots. can be mitigated through data science approaches combining artificial intelligence with statistics domain-based knowledge. In this paper, survey works aimed at characterization, feature extraction subsequent detection been conducted from perspective. Along it, an analysis 8 renowned repositories presented. Furthermore, case study on tweets related COVID-19 pandemic, factors behind during critical times, distinguishing between factual emotional viable restrain enunciated.

Язык: Английский

Процитировано

44
Review on the COVID-19 pandemic prevention and control system based on AI DOI
Junfei Yi, Hui Zhang, Jianxu Mao

и другие.

Engineering Applications of Artificial Intelligence, Год журнала: 2022, Номер 114, С. 105184 - 105184

Опубликована: Июль 11, 2022

Язык: Английский

Процитировано

37
DC-CNN: Dual-channel Convolutional Neural Networks with attention-pooling for fake news detection DOI Open Access
Kun Ma,

Changhao Tang,

Weijuan Zhang

и другие.

Applied Intelligence, Год журнала: 2022, Номер 53(7), С. 8354 - 8369

Опубликована: Авг. 1, 2022

Язык: Английский

Процитировано

34