A hybrid deep learning approach for detecting sentiment polarities and knowledge graph representation on monkeypox tweets

Decision Analytics Journal, Год журнала: 2023, Номер 7, С. 100243 - 100243

Опубликована: Май 4, 2023

People have recently begun communicating their thoughts and viewpoints through user-generated multimedia material on social networking websites. This information can be images, text, videos, or audio. With the help of knowledge graphs, it is possible to extract organized from texts images aid in semantic analysis. Recent years seen a rise frequency occurrence this pattern. Twitter one most extensively utilized media sites, also finest locations get sense how people feel about events that are linked Monkeypox sickness. because tweets shortened often updated, both which contribute platform's character. The fundamental objective study deeper comprehension diverse range reactions response presence condition. focuses determining what individuals think monkeypox illnesses, presenting hybrid technique based Convolutional Neural Networks (CNN) Long Short-Term Memory (LSTM). We considered all three polarities user's tweet: positive, negative, neutral. Knowledge graphs embedded various healthcare applications provide improved data representation inference, they been shown helpful analytics. describe graph related data, provides real-time eventful source new information. recommended model's accuracy was 94% tweet dataset. Other performance metrics such as accuracy, recall, F1-score were test our models results time resource-effective manner. findings then compared more traditional approaches machine learning. In addition, ability recognize has built into use graphs. research an increased awareness infection general population.

Язык: Английский

---

An Improved Ensemble Method for Predicting Hyperchloremia in Adults With Diabetic Ketoacidosis

IEEE Access, Год журнала: 2024, Номер 12, С. 9536 - 9549

Опубликована: Янв. 1, 2024

Diabetic ketoacidosis (DKA) is a serious complication that affects millions of individuals globally and presents significant health complications. Hyperchloremia, an electrolyte imbalance characterized by high levels chloride in the blood, may result gastrointestinal problems, kidney damage, even death, especially DKA patients. Early detection treatment hyperchloremia are utmost importance management DKA. This study explores potential bootstrap aggregating ensemble with random subspaces machine learning approach to predict occurrence hyperchloremia, providing basis for early intervention improved patient outcomes. We tested our retrospective MIMIC-III database containing 1177 patients compared it previous studies area under curve (AUC) 100%. Our showed performance outperforming other methods. The combination this enhance timely cases, ultimately leading outcomes more effective DKA-associated work aims contribute development decision support tools healthcare professionals, assisting them making informed decisions patients, focus on preventing managing hyperchloremia.

Язык: Английский

---

An Improved Framework for Detecting Thyroid Disease Using Filter-Based Feature Selection and Stacking Ensemble

IEEE Access, Год журнала: 2024, Номер 12, С. 89098 - 89112

Опубликована: Янв. 1, 2024

Язык: Английский

---

A natural language processing approach for analyzing COVID-19 vaccination response in multi-language and geo-localized tweets

Healthcare Analytics, Год журнала: 2023, Номер 3, С. 100172 - 100172

Опубликована: Апрель 11, 2023

Social media platforms, such as Twitter, have been paramount in the COVID-19 context due to their ability collect public concerns about vaccination campaign, which has underway end pandemic. This worldwide campaign heavily relied on actual willingness of individuals get vaccinated independently language they speak or country reside. study analyzes Twitter posts Pfizer/BioNTech, Moderna, AstraZeneca/Vaxzevria, and Johnson & vaccines by considering most spoken western languages. Tweets were sampled between April 15 September 15, 2022, after injections at least three doses, collecting 9,513,063 that contained vaccine-related keywords. To determine success vaccination, temporal sentiment analysis conducted, reporting opinion changes over time corresponding events whenever possible concerning each vaccine. Furthermore, we extracted main topics languages providing potential bias language-specific dictionary, Moderna Spanish, grouped them per country. Once performed pre-processed procedure worked with 8,343,490 tweets. Our findings show Pfizer debated vaccine worldwide, side effects pregnant women children heart diseases.

Язык: Английский

---

Off-label drug use during the COVID-19 pandemic in Africa: topic modelling and sentiment analysis of ivermectin in South Africa and Nigeria as a case study

Journal of The Royal Society Interface, Год журнала: 2023, Номер 20(206)

Опубликована: Сен. 1, 2023

Although rejected by the World Health Organization, human and even veterinary formulation of ivermectin has widely been used for prevention treatment COVID-19. In this work we leverage Twitter to understand reasons drug use from supporters, their source information, emotions, gender demographics, location in Nigeria South Africa. Topic modelling is performed on a dataset gathered using keywords ‘ivermectin’ ‘ivm’. A model fine-tuned RoBERTa find stance tweets. Statistical analysis compare emotions. Most supporters either redistribute conspiracy theories posted influencers, or refer flawed studies confirming efficacy vitro . Three emotions have highest intensity, optimism, joy disgust. The number anti-ivermectin tweets significant positive correlation with vaccination rate. All provinces Africa most are pro-ivermectin higher disgust polarity. This makes effort public discussions regarding during COVID-19 pandemic help policy-makers rationale behind its popularity, inform more targeted policies discourage self-administration ivermectin. Moreover, it lesson future outbreaks.

Язык: Английский

---

The Role of Social Media in Xenophobic Attack in South Africa

Опубликована: Фев. 16, 2024

Xenophobia is a pressing issue in South Africa, with frequent instances of violence against immigrants. With the rise social media, platforms like Twitter reflect public sentiment on this matter. This study examines tweets from 2017 to 2022 about xenophobia using NLP, analysis, and machine learning understand feelings predict potential xenophobic incidents. The findings aim help policymakers devise strategies enhance cohesion promote more inclusive society.

Язык: Английский

---

Analyzing the Perception of Immigrants in South Africa: A Machine Learning Approach to Aggregate Twitter Sentiment Data

2021 IST-Africa Conference (IST-Africa), Год журнала: 2024, Номер unknown, С. 1 - 8

Опубликована: Май 20, 2024

Язык: Английский

---

Evaluating automatic annotation of lexicon-based models for stance detection of M-pox tweets from May 1st to Sep 5th, 2022

PLOS Digital Health, Год журнала: 2024, Номер 3(7), С. e0000545 - e0000545

Опубликована: Июль 30, 2024

Manually labeling data for supervised learning is time and energy consuming; therefore, lexicon-based models such as VADER TextBlob are used to automatically label data. However, it argued that automated labels do not have the accuracy required training an efficient model. Although frequently stance detection, been properly evaluated, in previous works. In this work, assess of analysis, we first manually a Twitter, now X, dataset related M-pox detection. We then fine-tune different transformer-based on hand-labeled dataset, compare their before after fine-tuning, with labeled Our results indicated fine-tuned surpassed by up 38% 72.5%, respectively. Topic modeling further shows fine-tuning diminished scope misclassified tweets specific sub-topics. conclude transformer elevates superior level significantly higher than detection labels. This study verifies reliable sensitive use-cases health-related purposes. more convenient developing Natural Language Processing (NLP) analyze mass opinions conversations social media platforms, during crises pandemics epidemics.

Язык: Английский

---

A machine learning approach towards assessing consistency and reproducibility: an application to graft survival across three kidney transplantation eras

Frontiers in Digital Health, Год журнала: 2024, Номер 6

Опубликована: Сен. 3, 2024

Background In South Africa, between 1966 and 2014, there were three kidney transplant eras defined by evolving access to certain immunosuppressive therapies as Pre-CYA (before availability of cyclosporine), CYA (when cyclosporine became available), New-Gen (availability tacrolimus mycophenolic acid). As such, factors influencing graft failure may vary across these eras. Therefore, evaluating the consistency reproducibility models developed study variations using machine learning (ML) algorithms could enhance our understanding post-transplant survival dynamics Methods This explored effectiveness nine ML in predicting 10-year We internally validated data spanning specified The predictive performance was assessed area under curve (AUC) receiver operating characteristics (ROC), supported other evaluation metrics. employed local interpretable model-agnostic explanations provide detailed interpretations individual model predictions used permutation importance assess global feature each era. Results Overall, proportion decreased from 41.5% era 15.1% Our best-performing demonstrated high accuracy. Notably, ensemble models, particularly Extra Trees model, emerged standout performers, consistently achieving AUC scores 0.95, 0.97 indicates that achieved outcomes. Among features evaluated, recipient age donor only throughout eras, while such glomerular filtration rate ethnicity showed specific resulting relatively poor historical transportability best model. Conclusions emphasises significance analysing post-kidney outcomes identifying era-specific mitigating failure. proposed framework can serve a foundation for future research assist physicians patients at risk

Язык: Английский

---

Comprehensive analytics of COVID-19 vaccine research: From topic modeling to topic classification

Artificial Intelligence in Medicine, Год журнала: 2024, Номер 157, С. 102980 - 102980

Опубликована: Сен. 18, 2024

Язык: Английский

---