Sustainable development plays a vital role in information and communication technology. In times of pandemics such as COVID-19, vulnerable people need help to survive. This help includes the distribution of relief packages and materials by the government with the primary objective of lessening the economic and psychological effects on the citizens affected by disasters such as the COVID-19 pandemic. However, there has not been an efficient way to monitor public funds’ accountability and transparency, especially in developing countries such as Nigeria. The understanding of public emotions by the government on distributed palliatives is important as it would indicate the reach and impact of the distribution exercise. Although several studies on English emotion classification have been conducted, these studies are not portable to a wider inclusive Nigerian case. This is because Informal Nigerian English (Pidgin), which Nigerians widely speak, has quite a different vocabulary from Standard English, thus limiting the applicability of the emotion classification of Standard English machine learning models. An Informal Nigerian English (Pidgin English) emotions dataset is constructed, pre-processed, and annotated. The dataset is then used to classify five emotion classes (anger, sadness, joy, fear, and disgust) on the COVID-19 palliatives and relief aid distribution in Nigeria using standard machine learning (ML) algorithms. Six ML algorithms are used in this study, and a comparative analysis of their performance is conducted. The algorithms are Multinomial Naïve Bayes (MNB), Support Vector Machine (SVM), Random Forest (RF), Logistics Regression (LR), K-Nearest Neighbor (KNN), and Decision Tree (DT). The conducted experiments reveal that Support Vector Machine outperforms the remaining classifiers with the highest accuracy of 88%. The “disgust” emotion class surpassed other emotion classes, i.e., sadness, joy, fear, and anger, with the highest number of counts from the classification conducted on the constructed dataset. Additionally, the conducted correlation analysis shows a significant relationship between the emotion classes of “Joy” and “Fear”, which implies that the public is excited about the palliatives’ distribution but afraid of inequality and transparency in the distribution process due to reasons such as corruption. Conclusively, the results from this experiment clearly show that the public emotions on COVID-19 support and relief aid packages’ distribution in Nigeria were not satisfactory, considering that the negative emotions from the public outnumbered the public happiness.

Background: Large language models (LLMs), such as GPT-3.5 and GPT-4 (OpenAI), have been transforming virtual patient systems in medical education by providing scalable and cost-effective alternatives to standardized patients. However, systematic evaluations of their performance, particularly for multimorbidity scenarios involving multiple coexisting diseases, are still limited. Objective: This systematic review aimed to evaluate LLM-based virtual patient systems for medical history-taking, addressing four research questions: (1) simulated patient types and disease scope, (2) performance-enhancing techniques, (3) experimental designs and evaluation metrics, and (4) dataset characteristics and availability. Methods: Following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020, 9 databases were searched (January 1, 2020, to August 18, 2025). Nontransformer LLMs and non–history-taking tasks were excluded. Multidimensional quality and bias assessments were conducted. Results: A total of 39 studies were included, screened by one computer science researcher under supervision. LLM-based virtual patient systems mainly simulated internal medicine and mental health disorders, with many addressing distinct single disease types but few covering multimorbidity or rare conditions. Techniques like role-based prompts, few-shot learning, multiagent frameworks, knowledge graph (KG) integration (top-k accuracy 16.02%), and fine-tuning enhanced dialogue and diagnostic accuracy. Multimodal inputs (eg, speech and imaging) improved immersion and realism. Evaluations, typically involving 10-50 students and 3-10 experts, demonstrated strong performance (top-k accuracy: 0.45-0.98, hallucination rate: 0.31%–5%, System Usability Scale [SUS] ≥80). However, small samples, inconsistent metrics, and limited controls restricted generalizability. Common datasets such as MIMIC-III (Medical Information Mart for Intensive Care-III) exhibited intensive care unit (ICU) bias and lacked diversity, affecting reproducibility and external validity. Conclusions: Included studies showed moderate risk of bias, inconsistent metrics, small cohorts, and limited dataset transparency. LLM-based virtual patient systems excel in simulating multiple disease types but lack multimorbidity patient representation. KGs improve top-k accuracy and support structured disease representation and reasoning. Future research should prioritize hybrid KG-chain-of-thought architectures integrated with open-source KGs (eg, UMLS [Unified Medical Language System] and SNOMED-CT [Systematized Nomenclature of Medicine - Clinical Terms]), parameter-efficient fine-tuning, dialogue compression, multimodal LLMs, standardized metrics, larger cohorts, and open-access multimodal datasets to further enhance realism, diagnostic accuracy, fairness, and educational utility.

This study examines the viability of Zero Trust (ZT) models, a burgeoning measure of cybersecurity, in different cultural contexts. The ZT security model is based on the principle of "never trust, always verify" and, unlike traditional models, rejects inherent trust assumptions. The focus is on Middle Eastern culture, specifically the United Arab Emirates (UAE), to examine impact of information security, national and organizational culture, and how these correlates with the adoption of the ZT model. A theoretical model explains the user and organizational behavior towards ZT in the Arab culture, assessed through a survey based on the most important factors of the information security culture. The empirical data was analyzed in the UAE with the participation of 98 cybersecurity professionals from 98 different organizations using SmartPLS4 and PLS-SEM. The overall results indicate that both national and organizational culture, as well as information security culture, are significantly and positively correlated with the adoption of the Zero Trust Architecture (ZTA). This indicates that cultural aspects at different levels of national, organizational, and information security-specific play a crucial role in the decision-making process regarding the implementation of ZT models. The inclusion of the UAE brings a particular cultural element that has unique variations, and practical recommendations to help organizations improve their ZT adoption in line with cultural considerations. These findings are relevant for organizations and policy makers. The inclusion of additional factors, countries, and participants in future research could increase the accuracy of the results.

Detecting user affect automatically during real-time conversation is the main challenge towards our greater aim of infusing social intelligence into a natural-language mixed-initiative High-Fidelity (Hi-Fi) audio control spoken dialog agent. In recent years, studies on affect detection from voice have moved on to using realistic, non-acted data, which is subtler. However, it is more challenging to perceive subtler emotions and this is demonstrated in tasks such as labelling and machine prediction. This paper attempts to address part of this challenge by considering the role of user satisfaction ratings and also conversational/dialog features in discriminating contentment and frustration, two types of emotions that are known to be prevalent within spoken human-computer interaction. However, given the laboratory constraints, users might be positively biased when rating the system, indirectly making the reliability of the satisfaction data questionable. Machine learning experiments were conducted on two datasets, users and annotators, which were then compared in order to assess the reliability of these datasets. Our results indicated that standard classifiers were significantly more successful in discriminating the abovementioned emotions and their intensities (reflected by user satisfaction ratings) from annotator data than from user data. These results corroborated that: first, satisfaction data could be used directly as an alternative target variable to model affect, and that they could be predicted exclusively by dialog features. Second, these were only true when trying to predict the abovementioned emotions using annotator?s data, suggesting that user bias does exist in a laboratory-led evaluation