Advancements in Generative AI: A Comprehensive Review of GANs, GPT, Autoencoders, Diffusion Model, and Transformers

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 69812 - 69837

Published: Jan. 1, 2024

Language: Английский

---

Advances in natural language processing for healthcare: A comprehensive review of techniques, applications, and future directions

Computer Science Review, Journal Year: 2025, Volume and Issue: 56, P. 100725 - 100725

Published: Feb. 6, 2025

Language: Английский

---

Social media sentiment analysis and opinion mining in public security: Taxonomy, trend analysis, issues and future directions

Journal of King Saud University - Computer and Information Sciences, Journal Year: 2023, Volume and Issue: 35(9), P. 101776 - 101776

Published: Sept. 26, 2023

The interest in social media sentiment analysis and opinion mining for public security events has increased over the years. availability of platforms communication provides a valuable source information research. content shared across gives potential input to physical environment phenomena related threats. been used to: monitor threats or emergency events, analyzing opinionated data threat management detection using geographic location-based analysis. However, systematic survey that describes trends latest developments this domain is unavailable. This paper presents security. aims understand progress current state-of-the-art, identify research gaps, propose future directions. In total, 200 articles published from 2016 2023 were considered survey. taxonomy shows key attributes limitations work presented surveyed articles. Subsequently, direction on suggested interested researchers.

Language: Английский

---

Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 106733 - 106748

Published: Jan. 1, 2023

Distributed denial of service (DDoS) attacks continue to be a major security concern, threatening the availability and reliability network services. Software-defined networking (SDN) has emerged as promising solution address this issue, enabling centralized control management. However, conventional SDN-based DDoS mitigation techniques often struggle detect mitigate sophisticated due their limited ability analyze complex traffic patterns. This paper proposes an innovative optimized approach that effectively combines mininet, Ryu controller, one dimensional-convolutional neural (1D-CNN) in SDN environments. The proposed involves training 1D-CNN model with labeled data identify abnormal patterns associated attacks. Furthermore, seven hyperparameters trained were tuned using non-dominated sorting genetic algorithm II (NSGA-II) achieve best accuracy minimum time. Once detects attack, controller dynamically adapts policies employs appropriate protect infrastructure. To evaluate effectiveness model, extensive experiments conducted simulated environment realistic attack dataset. experimental results demonstrate developed achieves significantly improved detection 99.99% compared other machine learning (ML) models. NSGA-II enhances improvement rate 9.5%, 8%, 5.4%, 2.6% when it is logistic regression (LR), random forest (RF), support vector (SVM), k-nearest neighbor (KNN) models respectively. research paves way for future developments leveraging deep (DL) driven architectures evolving cybersecurity challenges.

Language: Английский

---

An Enhanced IDBO-CNN-BiLSTM Model for Sentiment Analysis of Natural Disaster Tweets

Biomimetics, Journal Year: 2024, Volume and Issue: 9(9), P. 533 - 533

Published: Sept. 4, 2024

The Internet's development has prompted social media to become an essential channel for disseminating disaster-related information. Increasing the accuracy of emotional polarity recognition in tweets is conducive government or rescue organizations understanding public's demands and responding appropriately. Existing sentiment analysis models have some limitations applicability. Therefore, this research proposes IDBO-CNN-BiLSTM model combining swarm intelligence optimization algorithm deep learning methods. First, Dung Beetle Optimization (DBO) improved by adopting Latin hypercube sampling, integrating Osprey Algorithm (OOA), introducing adaptive Gaussian-Cauchy mixture mutation disturbance. DBO (IDBO) then utilized optimize Convolutional Neural Network-Bidirectional Long Short-Term Memory (CNN-BiLSTM) model's hyperparameters. Finally, constructed classify tendencies associated with Hurricane Harvey event. empirical indicates that proposed achieves 0.8033, outperforming other single hybrid models. In contrast GWO, WOA, algorithms, enhanced 2.89%, 2.82%, 2.72%, respectively. This study proves can be applied assist emergency decision-making natural disasters.

Language: Английский

---

Enhancing Bitcoin Price Prediction with Deep Learning: Integrating Social Media Sentiment and Historical Data

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(3), P. 1554 - 1554

Published: Feb. 4, 2025

Bitcoin, the pioneering cryptocurrency, is renowned for its extreme volatility and speculative nature, making accurate price prediction a persistent challenge investors. While recent studies have employed multivariate models to integrate historical data with social media sentiment analysis, this study focuses on improving an existing univariate approach By incorporating tweet volume into framework, we systematically evaluated benefits of integration. Among five LSTM-based developed study, Multi-LSTM-Sentiment model achieved best performance, lowest mean absolute error (MAE) 0.00196 root-mean-square (RMSE) 0.00304. These results underscore significance including in predictive modelling demonstrate potential enhance decision-making highly dynamic cryptocurrency market.

Language: Английский

---

Review on Sentiment Analysis of Tweets Using Machine Learning Techniques: A Data Science Perspective

Lecture notes in networks and systems, Journal Year: 2025, Volume and Issue: unknown, P. 495 - 512

Published: Jan. 1, 2025

Language: Английский

---

Unpacking Online Discourse on Bioplastics: Insights from Reddit Sentiment Analysis

Polymers, Journal Year: 2025, Volume and Issue: 17(6), P. 823 - 823

Published: March 20, 2025

Bioplastics have been presented as a sustainable alternative to products derived from fossil sources. In response, industries developed innovative using biopolymers across various sectors, such food, packaging, biomedical, and construction. However, consumer acceptance remains crucial for their widespread adoption. This study aims explore public sentiment toward bioplastics, focusing on emotions expressed Reddit. A dataset of 5041 Reddit comments was collected keywords associated with bioplastics the extraction process facilitated by Python-based libraries like pandas, NLTK, NumPy. The analysis conducted NRCLex, broadly used lexicon. overall findings suggest that trust, anticipation, joy were most dominant in time frame 2014–2024, indicating emotional response towards has mostly positive. Negative fear, sadness, anger less prevalent, although an intense noted 2018. Findings also indicate temporal co-occurrence between significant events related changes among users. Although representativeness sample is limited, results this support need develop real-time monitoring public’s responses. Thus, it will be possible design communication campaigns more aligned needs.

Language: Английский

---

Combining large language and deep learning models for media text classification to perform supply chain due diligence risk assessments

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

Published: May 9, 2025

New legislation requires companies to regularly perform environmental, social, and governance (ESG) risk assessments for their suppliers. This is typically done using public indicators countries industries in which the suppliers operate. However, approach often does not represent actual stemming from a particular supplier accurately. Moreover, are usually only updated annually do reflect current developments. Therefore, big text data collected media monitoring on can augment these provide more accurate timely assessments. Although, texts this purpose challenging. It has low signal-to-noise ratio reliable complex interpretation be usable. Against background, we propose hybrid of large language models deep learning classifiers process automated ESG assessment. A set 17 investigated tested against direct usage classification. The performance evaluated Monte Carlo experiments four distinct datasets.

Language: Английский

---

Analyzing Public Reactions, Perceptions, and Attitudes during the MPox Outbreak: Findings from Topic Modeling of Tweets

Computers, Journal Year: 2023, Volume and Issue: 12(10), P. 191 - 191

Published: Sept. 23, 2023

In the last decade and a half, world has experienced outbreaks of range viruses such as COVID-19, H1N1, flu, Ebola, Zika virus, Middle East Respiratory Syndrome (MERS), measles, West Nile just to name few. During these virus outbreaks, usage effectiveness social media platforms increased significantly, served virtual communities, enabling their users share exchange information, news, perspectives, opinions, ideas, comments related outbreaks. Analysis this Big Data conversations using concepts Natural Language Processing Topic Modeling attracted attention researchers from different disciplines Healthcare, Epidemiology, Science, Medicine, Computer Science. The recent outbreak MPox resulted in tremendous increase Twitter. Prior works area research have primarily focused on sentiment analysis content Tweets, few that topic modeling multiple limitations. This paper aims address gap makes two scientific contributions field. First, it presents results performing 601,432 Tweets about 2022 Mpox were posted Twitter between 7 May 3 March 2023. indicate during time may be broadly categorized into four distinct themes—Views Perspectives Mpox, Updates Cases Investigations LGBTQIA+ Community, COVID-19. Second, findings Tweets. show theme was most popular (in terms number posted) Views Mpox. followed by which themes COVID-19 respectively. Finally, comparison with studies is also presented highlight novelty significance work.

Language: Английский

---