A Novel text2IMG Mechanism of Credit Card Fraud Detection: A Deep Learning Approach

Electronics, Journal Year: 2022, Volume and Issue: 11(5), P. 756 - 756

Published: March 1, 2022

Online sales and purchases are increasing daily, they generally involve credit card transactions. This not only provides convenience to the end-user but also increases frequency of online fraud. In recent years, in some countries, this fraud increase has led an exponential detection, which become increasingly important address security issue. Recent studies have proposed machine learning (ML)-based solutions for detecting fraudulent transactions, their detection scores still need improvement due imbalance classes any given dataset. Few approaches achieved exceptional results on different datasets. study, Kaggle dataset was used develop a deep (DL)-based approach solve text data problem. A novel text2IMG conversion technique is that generates small images. The images fed into CNN architecture with class weights using inverse method resolve DL ML were applied verify robustness validity system. An accuracy 99.87% by Coarse-KNN features CNN.

Language: Английский

---

E-Government 3.0: An AI Model to Use for Enhanced Local Democracies

Sustainability, Journal Year: 2023, Volume and Issue: 15(12), P. 9572 - 9572

Published: June 14, 2023

While e-government (referring here to the first generation of e-government) was just simple manner delivering public services via electronic means, e-gov 2.0 refers use social media and Web technologies in government operations service delivery. However, term ‘e-government 2.0’ is becoming less common as focus shifts towards broader digital transformation initiatives that may include AI technologies, among others, such blockchain, virtual reality, augmented reality. In this study, we present relatively new concept 3.0, which built upon principles but emerging (e.g., artificial intelligence) transform delivery improve governance. The study objective explore potential 3.0 enhance citizen participation, delivery, increase responsiveness compliance administrative systems relation citizens by integrating into using a background evolution over time. paper analyzes challenges faced municipalities responding petitions, are core application local democracies. author starts presenting an example e-petition system (as today) analyses anonymized data text corpus petitions directed one Romania municipalities. He will propose model able deal faster more accurately with increased number inputs, trying promote it who, for some reason, still reluctant implement their operations. conclusions suggest be effective on improving algorithms rather than solely ‘old’ technologies.

Language: Английский

---

Combating Covid-19 using machine learning and deep learning: Applications, challenges, and future perspectives

Array, Journal Year: 2022, Volume and Issue: 17, P. 100271 - 100271

Published: Dec. 10, 2022

COVID-19, a worldwide pandemic that has affected many people and thousands of individuals have died due to during the last two years. Due benefits Artificial Intelligence (AI) in X-ray image interpretation, sound analysis, diagnosis, patient monitoring, CT identification, it been further researched area medical science period COVID-19. This study assessed performance investigated different machine learning (ML), deep (DL), combinations various ML, DL, AI approaches employed recent studies with diverse data formats combat problems arisen COVID-19 pandemic. Finally, this shows comparison among stand-alone ML DL-based research works regarding issues AI-based works. After in-depth analysis comparison, responds proposed questions presents future directions context. review work will guide groups develop viable applications based on models, also healthcare institutes, researchers, governments by showing them how these techniques can ease process tackling

Language: Английский

---

Correction: Alabrah et al. Gulf Countries’ Citizens’ Acceptance of COVID-19 Vaccines—A Machine Learning Approach. Mathematics 2022, 10, 467

Mathematics, Journal Year: 2025, Volume and Issue: 13(7), P. 1047 - 1047

Published: March 24, 2025

Affiliation Revision [...]

Language: Английский

---

Clarifying Misunderstandings in COVID-19 Vaccine Sentiment and Stance Analysis and Their Implications for Vaccine Hesitancy: A Systematic Review

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: March 25, 2025

Background Advances in machine learning (ML) models have increased the capability of researchers to detect vaccine hesitancy social media using Natural Language Processing (NLP). A considerable volume research has identified persistence COVID-19 discourse shared on various platforms. Methods Our objective this study was conduct a systematic review employing sentiment analysis or stance detection towards vaccines and vaccination spread Twitter (officially known as X since 2023). Following registration PROSPERO international registry reviews, we searched papers published from 1 January 2020 31 December 2023 that used supervised assess through Twitter. We categorized studies according taxonomy five dimensions: tweet sample selection approach, self-reported type, classification typology, annotation codebook definitions, interpretation results. analyzed if report different trends than those by examining how is measured, whether efforts were made avoid measurement bias. Results found bias widely prevalent analyze toward vaccination. The reporting errors are sufficiently serious they hinder generalisability these understanding individual opinions communicate reluctance vaccinate against SARS-CoV-2. Conclusion Improving NLP methods crucial addressing knowledge gaps discourse.

Language: Английский

---

Attention-Enriched Mini-BERT Fake News Analyzer Using the Arabic Language

Future Internet, Journal Year: 2023, Volume and Issue: 15(2), P. 44 - 44

Published: Jan. 22, 2023

Internet use resulted in people becoming more reliant on social media. Social media have become the main source of fake news or rumors. They spread uncertainty each sector real world, whether politics, sports, celebrities’ lives—all are affected by uncontrolled behavior platforms. Intelligent methods used to control this various languages already been much discussed and frequently proposed researchers. However, Arabic grammar language a far complex crucial learn. Therefore, work fake-news-based datasets related studies is needed other The current study uses recently published dataset annotated experts. Further, Arabic-language-based embeddings given machine learning (ML) classifiers, trained minibidirectional encoder representations from transformers (BERT) obtain sentiments feed deep (DL) classifier. holdout validation schemes applied both ML classifiers mini-BERT-based neural classifiers. results show consistent improvement performance which outperformed increasing training data. A comparison with previous detection shown where greater improvement.

Language: Английский

---

A Feature-Based Approach for Sentiment Quantification Using Machine Learning

Electronics, Journal Year: 2022, Volume and Issue: 11(6), P. 846 - 846

Published: March 8, 2022

Sentiment analysis has been one of the most active research areas in past decade due to its vast applications. quantification, a new problem this field, extends sentiment from individual documents an aggregated collection documents. widely researched, but quantification drawn less attention despite offering greater potential enhance current business intelligence systems. In research, perform framework based on feature engineering is proposed exploit diverse sets such as sentiment, content, and part speech, well deep features including word2vec GloVe. Different machine learning algorithms, conventional, ensemble learners, approaches, have investigated standard datasets SemEval2016, SemEval2017, STS-Gold, Sanders. The empirical-based results reveal effectiveness process when applied algorithms. also that ensemble-based algorithm AdaBoost outperforms other conventional algorithms using combination sets. RNN, hand, shows optimal word embedding-based features. This help applications polling, trend analysis, automatic summarization, rumor or fake news detection.

Language: Английский

---

Comparison of Machine Learning Approaches for Detecting COVID-19-Lockdown-Related Discussions During Recovery and Lockdown Periods

Journal of Operations Intelligence, Journal Year: 2023, Volume and Issue: 1(1), P. 11 - 29

Published: Oct. 25, 2023

Ever since COVID-19 was declared a pandemic, governments around the world have implemented numerous phases of lockdown measures to curb spread virus. These tactics manifest themselves in form widespread fear and panic driven by social media discussions. Given that individuals hold diverse opinions about these during after their completion, positive negative lockdown-related discussions should be differentiated further understand major related issues make appropriate messaging policy choices future. We conduct sentiment analysis (SA) COVID-19-lockdown-related tweets using different machine learning (ML) classifiers then evaluate performance before synthetic minority oversampling technique (SMOTE). This research is performed five phases, starting with data collection followed pre-processing dataset, preparing dataset annotation, applying SMOTE ML classifiers. observe an improvement accuracy ( ) as confirmed Matthew correlation coefficient across most classifiers, except for k-nearest neighbour (KNN), whose Acc decreased from 0.82 0.59 MCC 0.544 0.279 applied. Despite potential some this cannot considered ultimate solution, especially other datasets. The study provides insights into need benchmark integration balancing approaches addition considering additional metrics, such MCC, binary classification problems, SA.

Language: Английский

---

Sentiment analysis of tweets and government translations: Assessing China’s post-COVID-19 landscape for signs of withering or booming

Global Media and China, Journal Year: 2023, Volume and Issue: 8(2), P. 213 - 233

Published: June 1, 2023

This article aims to gain insights into the prevailing public sentiment during policy relaxation period by examining whether post-COVID-19 landscape reflects signs of withering or booming conditions. Employing methods from natural language processing (NLP) and machine learning (ML), analysis reveals a predominance positive December 7, 2022 May 17, 2023, indicative an optimistic perspective potentially flourishing environment. A predictive model based on logistic regression emerges as notably effective tool for prediction, suggesting potential utility in predicting future health crises. comparison sentiments translations government aligns with previous research, revealing less favorable depiction translated texts compared source texts. Furthermore, commonality index, measure group consensus value, surpasses typical range, while certainty confidence, slightly falls below norm. These findings offer valuable considerations highlighting areas international communication understanding improvement.

Language: Английский

---

Political and Economic Patterns in COVID-19 News: From Lockdown to Vaccination

IEEE Access, Journal Year: 2022, Volume and Issue: 10, P. 40036 - 40050

Published: Jan. 1, 2022

The purpose of this study is to analyse COVID-19 related news published across different geographical places, in order gain insights reporting differences. pandemic had a major outbreak January 2020 and was followed by preventive measures, lockdown, finally the process vaccination. To date, more comprehensive analysis are missing, especially those which explain what aspects being reported newspapers inserted economies belonging political alignments. Since LDA often less coherent when there articles world about an event you look answers for specific queries. It because having semantically content. address challenge, we performed pooling based on information retrieval using TF-IDF score data processing step topic modeling with combination 1 6 ngrams. We used VADER sentiment analyzer analyze differences sentiments places. novelty at how media, providing comparison among countries economic contexts. Our findings suggest that alignment support Also, issues depend economy place where newspaper resides.

Language: Английский

---