Arabic Spam Tweets Classification: A Comprehensive Machine Learning Approach DOI Creative Commons

Wafa Hussain Hantom,

Atta Rahman

AI, Journal Year: 2024, Volume and Issue: 5(3), P. 1049 - 1065

Published: July 2, 2024

Nowadays, one of the most common problems faced by Twitter (also known as X) users, including individuals well organizations, is dealing with spam tweets. The problem continues to proliferate due increasing popularity and number users social media platforms. Due this overwhelming interest, spammers can post texts, images, videos containing suspicious links that be used spread viruses, rumors, negative marketing, sarcasm, potentially hack user’s information. Spam detection among hottest research areas in natural language processing (NLP) cybersecurity. Several studies have been conducted regard, but they mainly focus on English language. However, Arabic tweet still has a long way go, especially emphasizing diverse dialects other than modern standard (MSA), since, tweets, dialect seldom used. situation demands an automated, robust, efficient approach. To address issue, research, various machine learning deep models investigated detect tweets Arabic, Random Forest (RF), Support Vector Machine (SVM), Naive Bayes (NB) Long-Short Term Memory (LSTM). In we focused words meaning text. Upon several experiments, proposed produced promising results contrast previous approaches for same datasets. showed RF classifier achieved 96.78% LSTM 94.56%, followed SVM 82% accuracy. Further, terms F1-score, there improvement 21.38%, 19.16% 5.2% using RF, classifiers compared schemes dataset.

Language: Английский

Vision-Based Construction Safety Monitoring Utilizing Temporal Analysis to Reduce False Alarms DOI Creative Commons
Syed Farhan Alam Zaidi, Jaehun Yang, Muhammad Sibtain Abbas

et al.

Buildings, Journal Year: 2024, Volume and Issue: 14(6), P. 1878 - 1878

Published: June 20, 2024

Construction safety requires real-time monitoring due to its hazardous nature. Existing vision-based systems classify each frame identify safe or unsafe scenes, often triggering false alarms object misdetection detection, which reduces the overall system’s performance. To overcome this problem, research introduces a system that leverages novel temporal-analysis-based algorithm reduce alarms. The proposed comprises three main modules: rule compliance, and temporal analysis. employs coordination correlation technique verify personal protective equipment (PPE), even with partially visible workers, overcoming common challenge on job sites. temporal-analysis module is key component evaluates multiple frames within time window, when hazard threshold exceeded, thus reducing experimental results demonstrate 95% accuracy an F1-score in scene classification, notable 2.03% average decrease during across five test videos. This study advances knowledge by introducing validating algorithm. approach not only improves reliability of safety-rule-compliance checks but also addresses challenges alarms, thereby enhancing management protocols environments.

Language: Английский

Citations

5
MEAG-YOLO: A Novel Approach for the Accurate Detection of Personal Protective Equipment in Substations DOI Creative Commons
Hong Zhang, Chunyang Mu,

Xing Ma

et al.

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(11), P. 4766 - 4766

Published: May 31, 2024

Timely and accurately detecting personal protective equipment (PPE) usage among workers is essential for substation safety management. However, traditional algorithms encounter difficulties in substations due to issues such as varying target scales, intricate backgrounds, many model parameters. Therefore, this paper proposes MEAG-YOLO, an enhanced PPE detection built upon YOLOv8n. First, the incorporates Multi-Scale Channel Attention (MSCA) module improve feature extraction. Second, it newly designs EC2f structure with one-dimensional convolution enhance fusion efficiency. Additionally, study optimizes Path Aggregation Network (PANet) learning of multi-scale targets. Finally, GhostConv integrated optimize operations reduce computational complexity. The experimental results show that MEAG-YOLO achieves a 2.4% increase precision compared YOLOv8n, 7.3% reduction FLOPs. These findings suggest effective identifying complex scenarios, contributing development smart grid systems.

Language: Английский

Citations

5
Detection of Labor Protection Wear on Jobsite Based on Improved Yolo Algorithm DOI
Bo Xu

Published: Jan. 1, 2025

Language: Английский

Citations

0
Artificial intelligence (AI) use for personal protective equipment training, remediation and education in healthcare DOI Creative Commons
Veronica Preda, Z. Ong, Chandana Wijeweera

et al.

American Journal of Infection Control, Journal Year: 2025, Volume and Issue: unknown

Published: March 1, 2025

Personal protective equipment (PPE) is a first-line transmission-based precaution for reducing the spread of nosocomial infections between healthcare workers (HCWs), patients, and staff. Disproportionate rates in HCWs during COVID-19 pandemic highlighted problematic skill gap effective PPE donning/doffing. We performed single-centre, mixed-method, prospective cohort study 293 Sydney, Australia. Participants were assessed using SXR AI-PPE®, an AI system that autonomously evaluates donning/doffing while providing real-time feedback on user technique. Quantitative data performance AI-guided unguided sessions recorded, including accuracy (%), time (sec) to don/doff, over multiple attempts. Additionally, questionnaires administered before after training assess changes self-efficacy correct use. Longitudinal results showed improved each guided session conducted at 3-monthly intervals, with 100% rate use two sessions. Following AI-PPE training, taken don doff was reduced by 15 22 seconds, respectively. These improvements maintained The AI-PPE® platform comprehensive tool capable real-time. platforms can effectively improve skills self-efficacy, implications contamination risk infections.

Language: Английский

Citations

0
TMU-GAN: a compliance detection algorithm for protective equipment in power operations DOI

Xuecun Yang,

Jiayu Li, Qingyun Zhang

et al.

Multimedia Systems, Journal Year: 2025, Volume and Issue: 31(3)

Published: May 2, 2025

Language: Английский

Citations

0
Intelligent Compliance Monitoring Using AI for Enhanced Staff Adherence in Retail DOI
Phuong Anh Nguyen,

Khang Minh Vuong,

Cuong Nhat Nguyen

et al.

IGI Global eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 195 - 210

Published: May 9, 2025

This paper explores the application of artificial intelligence (AI) in automating compliance monitoring through CCTV infrastructure, specifically focusing on tracking employee adherence to uniform policies and safety protocols. Despite growing use AI surveillance, few systems have been designed monitor staff behavior real-time ensure compliance. research addresses this gap by developing an AI-powered framework that leverages advanced computer vision techniques for human detection, re-identification, face verification. By implementing deep learning models, system accurately detects non-compliance events related identification standards, thereby reducing need manual oversight. The results demonstrate can significantly enhance efficiency reliability retail environments, with substantial implications operational risks improving workforce management.

Language: Английский

Citations

0
Comparison of Machine Learning Approaches for Robust and Timely Detection of PPE in Construction Sites DOI Creative Commons

rohollah Azizi,

Maria Koskinopoulou, Yvan Pétillot

et al.

Robotics, Journal Year: 2024, Volume and Issue: 13(2), P. 31 - 31

Published: Feb. 16, 2024

Globally, workplace safety is a critical concern, and statistics highlight the widespread impact of occupational hazards. According to International Labour Organization (ILO), an estimated 2.78 million work-related fatalities occur worldwide each year, with additional 374 non-fatal injuries illnesses. These incidents result in significant economic social costs, emphasizing urgent need for effective measures across industries. The construction sector particular faces substantial challenges, contributing notable share these due nature its operations. As technology, including machine vision algorithms robotics, continues advance, there growing opportunity enhance global standards mitigate human toll hazards on broader scale. This paper explores development evaluation two distinct designed accurate detection equipment sites. first algorithm leverages Faster R-CNN architecture, employing ResNet-50 as backbone robust object detection. Subsequently, results obtained from are compared those second algorithm, Few-Shot Object Detection (FsDet). selection FsDet motivated by efficiency addressing time-intensive process compiling datasets network training recognition. research methodology involves fine-tuning both assess their performance Comparative analysis aims evaluate effectiveness novel methods employed algorithms.

Language: Английский

Citations

3
Deep learning-based object detection for dynamic construction site management DOI
Jiayi Xu, Wei Pan

Automation in Construction, Journal Year: 2024, Volume and Issue: 165, P. 105494 - 105494

Published: June 15, 2024

Language: Английский

Citations

3
Advancing Food Safety Behavior with AI: Innovations and Opportunities in the Food Manufacturing Sector DOI Creative Commons
Ke Wang, Miranda Mirosa, Yakun Hou

et al.

Trends in Food Science & Technology, Journal Year: 2025, Volume and Issue: unknown, P. 105050 - 105050

Published: April 1, 2025

Language: Английский

Citations

0
Development and Optimization of a Construction Personal Protective Equipment (PPE) Detection Model on YOLOv8 Architecture DOI

Zidan Rafindra Utomo,

Prajanto Wahyu Adi, Priyo Sidik Sasongko

et al.

JURNAL MASYARAKAT INFORMATIKA, Journal Year: 2025, Volume and Issue: 16(1), P. 1 - 14

Published: April 28, 2025

Workplace safety in the construction sector remains a critical issue due to frequent accidents caused by non-compliance with Personal Protective Equipment (PPE) regulations. Manual supervision is inefficient and prone errors, necessitating an automated detection approach. The prior YOLOv5 version trained on Construction Safety dataset from Roboflow-100, achieves mean Average Precision ([email protected]) of 0.867. However, class imbalance, particularly underrepresentation "no-helmet" "no-vest" categories, limited performance. This study improves model tuning hyperparameters for optimal training using grid search applying data augmentation techniques address imbalance. Mosaic Mixup technique applied dataset. augmented used retrain YOLOv8, further optimizing accuracy. Results indicate improved [email protected] 0.921, demonstrating enhanced performance PPE violation detection. These refinements aim strengthen workplace enforcement through more accurate balanced

Language: Английский

Citations

0