نشریه پژوهش های نوین در شهر هوشمند
سال سوم شماره 1 (پاییز 1403)

Since the inception and establishment of cities, they have served as the foundation and hub for economic-social, political, and physical-spatial transformations. Throughout history, humanity has undergone two major revolutions: the agricultural revolution and the industrial revolution, and is currently experiencing the third revolution—the information revolution—which continues to impact cities and the way citizens live. This, in turn, leads to changes in the cities and the components that constitute them. The information revolution is increasingly influenced by new technologies. The metaverse, one of these new technologies with great potential, is considered as the future of technology and a possible replacement for the future internet. It is emerging and expanding within cities. The metaverse, as a combination of virtual and physical realities, merges the virtual world with the real one, holding high potential for urban planning and development. In the context of smart urban growth and smart cities, cities that are electronic, virtual, and digital are continuously evolving. The newest form of these cities can be referred to as metaverse cities, which, with the expansion of the metaverse, are experiencing both theoretical and practical developments. This study explores metaverse cities and the impact of the metaverse on cities regarding four key components of smart cities, including smart citizens as the main stakeholders and actors in the city, smart transportation as a crucial element of the city, smart tourism as an important part of the contemporary urban economy, and smart security as a vital aspect of citizens' lives. This research is a library-based study utilizing a descriptive-analytical approach, drawing on updated and credible sources. The overall results of the study indicate that the metaverse is increasingly expanding in today's world, both in terms of academic studies and practical implementation. Future smart cities will essentially be smart and metaverse cities.

Customer loyalty has always been one of the main concerns and challenges for domestic banks, as their lifecycle depends on customer loyalty. Various studies have been conducted on customer loyalty in banks and financial and commercial institutions. This research evaluated the use of data mining and five machine learning algorithms—Support Vector Machine, Naive Bayes, MLP, Decision Tree, and Logistic Regression—for detecting customer loyalty in a domestic bank. Three performance metrics—accuracy, precision, and recall—were used to compare the results. The data used in this research were collected from a domestic bank and consisted of 15 criteria for evaluating customer loyalty to the organization. The results showed that, among the five algorithms, the best performance was achieved by the Decision Tree, while the Logistic Regression algorithm showed the worst performance. Finally, to improve the results, a combination of a neural network and the Teacher-Student optimization algorithm was used to achieve better detection of loyal customers.

Global cyberattacks significantly impact the economy, society, organizations, and individuals. Existing research on cyberattacks, particularly in providing AI-based analytical solutions to share information on cyber threats at the national level, is limited. National cybersecurity strategists require AI-based decision support systems to assess the cybersecurity posture or preparedness of a country. This paper proposes an AI-based solution that autonomously collects multidimensional data on cyber-related incidents from social media posts. The proposed system offers crucial analytical capabilities across a spectrum of cyber threats, utilizing sophisticated AI algorithms for anomaly detection, prediction, sentiment analysis, location detection, translation, and more. The system has been operational from April 21, 2021, to May 31, 2023. In 21 days, the system independently collected 30,203 records on cyber threats, covering various aspects of cyber threats. These dimensions included daily records of cyberattacks nationwide, such as ransomware, exploits, web threats, spam, malicious emails, network attacks, local contamination, and demandbased scanning. Additionally, the system obtained and analyzed 3,789 cyber-related tweets from 3,402 users in 37 different languages using AI. It also independently translated 893 non-English tweets. The proposed system is the first solution to employ Convolutional Neural Networks (CNN) for anomaly detection in the global cyber threat spectrum and for the automatic prediction of cyberattacks. The system was demonstrated to provide evidence-based decisions on global cyber threats across multiple platforms, including iOS, Android, and Windows.

One of the key goals of universities is to prepare future managers for technologies that optimize various and diverse activities. It is evident that in a rapidly changing world, the future of graduates is increasingly dependent on AI-mediated work environments. The purpose of this research is to classify and describe the achievements of artificial intelligence in the higher education system. This research used a documentary and descriptive-analytical approach, studying available printed and electronic sources, including books, articles, and related websites, to examine the applications of artificial intelligence in higher education. The results of this study indicate that in the era of information and communication technology, social changes are as important as technological transformations, and AI is not just a subject for technological innovation but represents a fundamental change in the relationship between higher education and socio-economic benefits. Therefore, universities must establish strong policies and research programs to advance toward AI. Additionally, jobs and working methods may undergo fundamental changes. As a result, it is expected that universities will go beyond significant digital transformation to meet the needs of today’s generation by analyzing student performance, predicting student dropouts and employment outcomes, improving administrative performance, and ultimately evaluating university services.

This study delves into the transformative role of deep learning and neural networks in the domain of air pollution control. By focusing on enhanced detection and monitoring, particularly through convolutional and recurrent neural architectures, the research highlights the potential of these technologies to unravel complex patterns within air quality dynamics. Beyond mere detection, these models demonstrate proactive capabilities, enabling the prediction and forecasting of pollution events. This foresight empowers the implementation of adaptive control strategies, effectively minimizing health risks and optimizing resource allocation. However, the study acknowledges challenges related to data quality and interpretability, emphasizing the necessity for interdisciplinary collaboration among machine learning experts, environmental scientists, and policymakers. In synthesizing these findings, the research contributes to the advancement of sustainable strategies for mitigating the impact of air pollution on human health and the environment and also reviews methods of controlling it by deep learning approaches.

The increase in motor vehicles has greatly increased air pollution in the Tehran metropolis and turned it into a challenge. By analyzing transportation pollutants, low-carbon cities and sustainable development strategies for urban districts can be achieved. This study deals with the spatiotemporal analysis of pollutants and transportation accessibility in the urban districts of Tehran. The concentrations of CO and NO2 pollutants were extracted from the Sentinel-5P satellite using coding in the GEE environment. OSM data were used to calculate the accessibility of the 22 urban districts. Moran's global autocorrelation indicated a clustered spatial distribution pattern of pollutants. The high autocorrelation value equal to 0.94 for CO and NO2 pollutants indicated their cumulative state in the atmosphere. By performing LISA, most clusters of these pollutants were identified in the central districts and southern suburbs. The average accessibility indicator of the 22 districts was 24.82 km/km2. About half of Tehran's urban districts had a high density of road transport networks. These districts are located in CO and NO2 clusters owing to the high traffic volume of motor vehicles. The results of this study emphasize the factors that aggravate Tehran's air pollution, including the clustered pattern of pollutants in the atmosphere and dense transportation network in urban districts. By integrating air pollution analysis and accessibility estimations into urban planning frameworks, cities can strive to achieve long-term sustainability goals.