Journal of Internet Services and Applications

Training Neural Networks in Cloud Environments: A Methodology and a Comparative Analysis

2025-04-14T17:10:25+00:00

Deep neural networks are solutions to problems that involve pattern recognition, and various works seek to optimize the performance of these networks. This optimization requires suitable hardware, which can be expensive for small and medium organizations. This work proposes a methodology to evaluate the performance and cost of training deep neural networks by assessing how much impact factors such as environment setup, frameworks, and datasets can have on training time and, along with this task, evaluating the total financial cost of the environment for the training process. Experiments were performed to measure and compare the performance and cost of training deep neural networks on cloud platforms such as Azure, AWS, and Google Cloud. In this sense, factors such as the size of the input image and the network architecture significantly impact the training time metric and the total cost.

How Does Software Configuration Parameters Impact Job's Execution Time in Spark?

2025-03-11T16:15:31+00:00

Traditional centralized systems cannot deal with the big data context. Distributed computing platforms such as Apache Spark have been widely adopted, but configuring their parameters is challenging given the number of factors and their interactions. This work employs Design of Experiments (DoE) techniques to screening most relevant factors regarding execution time of a Naïve Bayes machine learning distributed task on a subset of the PT7 Web Corpus, which has 14.88 GB of data. Employing a fractional factorial design with 192 experimental units and linear regression techniques with backward elimination, we obtained (i) the most relevant factors based on statistical significance and (ii) a model capable of predicting execution time according to parameters' values in the analyzed context. Our results also include a visualization technique based on Inselberg's Parallel Coordinates to comprehend the impact on performance facing various configuration possibilities.

Syntactic and Semantic Edge Interoperability

2025-02-03T13:26:26+00:00

The Internet of Things (IoT) has transformed various sectors, from home automation to healthcare, leveraging a multitude of sensors and actuators communicating through cloud, fog, and edge networks. However, the diversity in device manufacturing and communication protocols necessitates interoperable communication interfaces. Most existing IoT interoperability solutions often rely on cloud-based centralised architectures and suffer from latency and scalability issues. This work specifically focuses on scenarios where decisions need to be made with IoT edge devices in real-time, even in situations where there might be internet disruptions, low bandwidth, or no internet connection. While typical IoT interoperability solutions support edge devices, their reliance on cloud-based architectures makes them unsuitable for mission-critical applications, environmental monitoring, or water quality monitoring, where internet connectivity cannot be guaranteed. To tackle these challenges, the project InterEdge proposed a theoretical interoperability model supporting hierarchical decentralised communication between edge devices. The aforementioned framework has four levels to handle network, syntactic, semantic, and organisational aspects of interoperability. As part of the same project, this work focuses on the implementation of the syntactic and semantic levels of the aforementioned framework. This work involves tackling the implementation challenges, particularly considering key issues related to transmission latency and memory requirements. We have created profiles for edge devices and data formats to store their essential and extra information. Using the profiles, communications can be established and maintained seamlessly among edge devices. We have conducted a comparative analysis between InterEdge implementation and three other implementations of established open standards. The experimental results demonstrate that the syntactic and semantic levels of the implemented interoperability solution, InterEdge, significantly outperforms the existing open standards in terms of standard benchmarking metrics such as code size, memory usage, and response latency. The contribution of this paper lies in these implementation results, which provide concrete evidence of the superior performance of our proposed solution, InterEdge, thereby validating its efficacy in real-world IoT scenarios.

Forwarding Metrology with an IoT and Blockchain Approach: The Gas Pumps Use Case

2025-03-16T13:07:26+00:00

Volumetric fraud at gas pumps is a serious and ongoing issue, leading to substantial financial losses for consumers.
Recognizing the severity of this problem, regulatory agencies in Brazil have introduced new gas pumps equipped with digital certification. This initiative is part of a broader strategy to integrate digital certification into measuring instruments, starting with gas pumps, as a proactive measure to counteract fraud. Additionally, Brazilian agencies are developing a mobile application that will allow users to access refueling data and perform their own inspections. In this context, we aim to further advance the topic by proposing a system that enhances the smart capabilities of metrology. We introduce the adoption of blockchain technology to establish consumer communities and promote metrology within IoT and cloud computing landscape. Our proposal allows users to share their refueling data, facilitating more active gas pump inspections with less dependence on regulatory agencies. We propose utilizing the user data in an evaluation system that cross-references the data and applies statistical methods to detect both volumetric fraud and fuel tampering. The system is designed to generate a ranking of gas pumps with the highest potential of being fraudulent, providing insights to both regulatory agencies and consumers. By leveraging blockchain technology, we can securely record user data and deliver the evaluation service in a transparent and decentralized manner. Finally, we apply statistical techniques to address the issue of trusting external data incorporated into the blockchain, which is commonly referred to as the oracle problem in the literature. Our simulation results demonstrate that, by using only refueling and vehicle data, we can achieve fraud detection accuracy of 89% in scenarios that closely resemble real-world conditions.

Intrusion detection in vehicular networks using machine learning

2025-03-16T13:16:00+00:00

Vehicular networks and intelligent transport systems play a critical role in modern urban mobility. In order to improve urban transportation in smart cities, vehicles and fixed stations exchange information about traffic, road conditions, and accidents, allowing better decision-making and ensuring greater safety for the population. However, to provide security, a vehicular network must be resilient to attacks. Anomaly detection models are a potential solution to the reduced effectiveness of signature-based intrusion detection systems, which struggle to detect new attacks due to the absence of previous signatures. Leveraging artificial intelligence in intrusion detection systems becomes relevant, as it allows learning from a vast amount of data. However, many models proposed for anomaly detection based on machine learning lack validation and application in vehicular networks, thus lacking evidence of promising results in these specific contexts. Therefore, this work aims to address this gap by comparing two models used in anomaly detection in the context of vehicular networks: the CNN-LSTM model that has already been applied in the area of vehicular networks and the TranAD model that needed to be adapted for this type of network. The results demonstrate that the CNN-LSTM model provides superior performance, presenting an F1 of 0.9585 against 0.8839 of TranAD in the scenario in which both models obtained the best result.

Spatial Optimization of Charging Networks for Heavy-Duty EVs Using Hexagonal Discrete Models

2025-02-12T12:20:02+00:00

Due to the environmental impact caused by greenhouse gas emissions, solving problems aimed at increasing the usage of electric vehicles became important. Personal Electric Vehicles are being highly adopted by society in order to reduce emissions. However, a prominent part of air pollution is provided by heavy-duty vehicles, such as trucks, and its electrification is challenging because of the lack of government policies and charging infrastructure. In light of this, electric charge stations should be located considering the truck drivers’ route to increase its adoption. Therefore, this study proposes two hexagonal discrete covering models, a Hexagonal P-Median (HPMP) and Hexagonal Capacitated Location Set Covering (HCLSCP), enhancing the space complexity of classical discrete models to cover the Brazilian truck drivers’ route. Furthermore, we compare the novel hexagonal models to a greedy method using a spatio-temporal simulation. We consider the infrastructure limitations with capacity constraints and waiting time in recharging queues with real-world data comprising locations of 3,086 drivers. The results show a trade-off between infrastructure cost, coverage demand, and queuing performance. The HPMP is ideal for covering demand, while the greedy method minimizes infrastructure cost, and HCLSCP outperforms the other models in queuing management.

Blockchain-Based location validation in an environment of mutual distrust

2025-04-14T17:26:08+00:00

As technology advances, more and more of our day-to-day objects are made smart and communicate with each other. Vehicles are no exception. They can exchange data to increase safety on the road or find more efficient routes, among other conveniences. Just like every other moving object, many of these applications are location-dependent. However, the network is vulnerable to safety and privacy threats. Certain dishonest and misbehaving peers may try to exploit the network. Therefore, establishing trust between vehicles is one of the network's most important challenges. We propose a decentralized and verifiable model for verifying GPS data using Cartesi Rollups. In this way, we count on the scalability of the multilayer blockchain solution and avoid the communication overhead of a voting system. We evaluated this model by simulating it with SUMO, varying the density of fraudulent nodes and the signal range. Under these conditions, it was able to identify fraud attempts with more than 85% accuracy in the worst cases.

An Alternative Particle Filter-Driven ADR for Mobile Devices in LoRaWAN Networks

2025-02-07T21:15:18+00:00

LoRaWAN is a leading LPWAN technology for Internet of Things applications, known for its long-range communication and low-power consumption. Its ADR mechanism optimizes performance by adjusting transmission parameters, such as spreading factor and transmission power, based on network conditions. However, ADR faces significant limitations in environments with mobile end devices, where fluctuating signal quality leads to increased packet loss, inefficient energy usage, and reduced communication reliability. To address these challenges, PF-ADR, an alternative ADR scheme, is proposed for LoRaWAN networks with mobile devices. PF-ADR employs a particle filter to estimate a representative SNR value, maintaining multiple hypotheses of the communication channel state to enable more precise parameter adjustments. Simulations conducted under various scalability conditions reveal that PF-ADR achieves up to 29.5% higher packet delivery ratio compared to M-ADR, and 52.17% more than the standard ADR, while demonstrating a 43.19% improvement in energy efficiency over MB-ADR. Additionally, the algorithm reduces packet loss due to signal degradation while maintaining scalable performance in large networks. These results highlight the potential of PF-ADR to enhance communication reliability and energy efficiency in dynamic mobile environments.

Plant Disease Detection Using Federated Learning and Cloud Infrastructure for Scalability and Data Privacy

2025-05-14T16:52:10+00:00

Agriculture faces significant challenges from crop diseases, making early and accurate detection critical. Federated Learning (FL), an advancement in artificial intelligence (AI) and machine learning (ML), presents a promising solution by enabling collaborative model training on decentralized data without the need to share sensitive information. This article examines the application of FL in detecting plant diseases through image analysis, highlighting the role of cloud computing in addressing challenges related to data processing, storage, and model scalability. By leveraging decentralized data stored and processed in the cloud, FL develops robust models that not only improve detection accuracy but also generalize effectively to new data, promoting knowledge sharing while ensuring data privacy. The integration of cloud infrastructure enables FL to scale, providing resilience and productivity gains in agricultural practices. The results show that the proposed approach achieves a 99.71% accuracy using the VGG16 model after Federated Learning aggregation, while preserving data confidentiality, enhancing agricultural resilience, and benefiting from the scalability and flexibility offered by cloud computing.

Performance Evaluation of a Camera Surveillance System in Smart Buildings Using Queuing Models

2025-05-12T01:00:35+00:00

Security is increasingly prioritized, driving the use of camera surveillance in various settings such as companies, schools, and hospitals. Cameras deter crime and enable continuous monitoring. Integrating Edge and Fog Computing into these systems decentralizes data processing, allowing for faster responses to critical events. Challenges in deploying such systems include high costs, complex technology integration, and precise sizing. Costs cover cameras, Edge devices, cabling, and software, while integration requires technical expertise and time. Accurate sizing is essential to prevent resource under- or over-utilization. Analytical modeling helps simulate scenarios and calculate needed resources. This work proposes an M/M/c/K queuing model to assess surveillance system performance in smart buildings, considering data arrival rates and Edge and Fog container capacities. The model allows parameter customization to analyze various scenarios. Results show that increasing the number of containers more significantly improves system performance than increasing the number of cores per container.

Empowering Client Selection with Local Knowledge Distillation for Efficient Federated Learning in Non-IID Data

2025-05-16T20:43:14+00:00

Federated Learning (FL) is a distributed approach in which multiple devices collaborate to train a shared, global model (GM). During its training, client devices must frequently communicate their gradients to the central server to update the GM weights. This incurs significant communication costs (bandwidth utilization and the number of messages exchanged). The heterogeneous nature of clients’ local datasets poses an extra challenge to the model training. In this sense, we introduce FedSeleKDistill, Federated Selection and Knowledge Distillation Algorithm, to decrease the overall communication costs. FedSeleKDistill is an innovative combination of: (i) client selection, and (ii) knowledge distillation approaches with three main objectives: (i) reducing the number of devices training at every round; (ii) decreasing the number of rounds until convergence; and (iii) mitigating the effect of client’s heterogeneous data on the GM effectiveness. In this paper, we extend the results obtained from the initial paper presenting FedSeleKDistill. The additional experimental evaluations on the MNIST and German Traffic Signs Benchmark datasets demonstrate that FedSeleKDistill is highly efficient in training the GM until convergence in heterogeneous FL. FedSeleKDistill reaches a higher accuracy score and faster convergence than state-of-the-art models. Our results also show higher performance when analyzing the accuracy scores on the clients’ local datasets.

Urban Cultural Signature with Web Data: A Case Study with Google Places Venues

2025-03-24T15:47:18+00:00

Providing knowledge about the characteristics of diverse cultural groups worldwide and identifying cultural similarities between their respective occupation regions can yield significant economic and social benefits. However, much of the existing research in this field relies on user behavior data, which may limit scalability and generalization due to the difficulty in obtaining such data. To address this, our work focuses on extracting venue data from Google Places and proposing a methodology based on the Scenes concept to enrich this dataset for generating cultural signatures of urban areas. This approach also considers the influence of different area sizes. Using Curitiba, Brazil, and Chicago, USA, as case studies, the results demonstrate that the proposed method can identify cultural similarities between regions while supporting an area-division strategy for analyzing cities across different countries. The findings show consistency, as evidenced by the segmentation of Curitiba and Chicago into culturally distinct clusters. This highlights the societal benefits of the proposal, such as location recommendations based on cultural criteria and real-time service validation.

People Counting Application with Crowded Scenarios: A Case Study with TV Boxes as Edge Devices

2025-05-06T14:49:22+00:00

Counting people in various urban spaces using artificial intelligence enables a wide range of smart city applications, enhancing governance and improving citizens' quality of life. However, the rapid expansion of edge computing for these applications raises concerns about the growing volume of electronic waste. To address this challenge, our previous work demonstrated the feasibility of repurposing confiscated illegal TV boxes as Internet of Things (IoT) edge devices for machine learning applications, specifically for people counting using images captured by cameras. Despite promising results, experiments in crowded scenarios revealed a high Mean Absolute Error (MAE). In this work, we propose a patching technique applied to YOLOv8 models to mitigate this limitation. By employing this technique, we successfully reduced the MAE from 8.77 to 3.77 using the nano version of YOLOv8, converted to TensorFlow Lite, on a custom dataset collected at the entrance of a university restaurant. This work presents an effective solution for resource-constrained devices and promotes a sustainable approach to repurposing hardware that would otherwise contribute to electronic waste.

A Fuzzy Inference System for DDoS Identification in Fog Computing based on Energy Consumption

2025-04-29T14:02:59+00:00

Internet of Things (IoT) networks, characterized by their heterogeneous devices, standards, and features, along with limited energy resources, are particularly vulnerable to security threats. Fog computing, which processes data closer to the network edge (i.e., IoT devices), has emerged as a key paradigm for addressing these issues. The Message Queuing Telemetry Transport (MQTT) protocol is commonly used for communication between IoT and fog devices due to its simplicity and ease of implementation. However, MQTT does not include built-in security measures, making it susceptible to Distributed Denial of Service (DDoS) attacks. This paper identifies the main DDoS threats in the context of the MQTT protocol and proposes a Fuzzy Inference System (FIS) designed to detect and classify specific DDoS attack types. By analyzing energy consumption patterns in fog nodes, fuzzy logic infers the degree of membership of a DDoS attack in a fog node, providing a robust method for threat detection in IoT environments.

Which One is Better? Distributed Artificial Intelligence Strategies for Accurate Vehicular Emissions Forecasting

2025-05-16T20:45:33+00:00

The rapid expansion of urban vehicular networks has led to increasing carbon emissions, posing significant environmental challenges in densely populated areas. Accurate emission predictions are crucial for sustainable urban planning, but current methods face limitations in handling large-scale dynamic data while balancing latency, privacy, and communication efficiency. This paper proposes a comprehensive framework study that compares centralized, federated, and shared learning approaches for CO₂ emissions prediction in vehicular networks, using data from vehicles and roadside units (RSUs) to predict emissions in diverse urban scenarios. By evaluating the performance of each approach on latency, communication overhead, and prediction accuracy, this work provides insights into optimizing learning strategies for real-time, scalable, and privacy-preserving emissions management in intelligent transportation systems. The findings offer valuable guidance to urban planners and policymakers, fostering the development of sustainable urban mobility solutions.

Synthetic Driving Conditions Data Generation Using Federated Generative Adversarial Networks

2025-04-07T16:46:36+00:00

Road safety remains a global challenge, especially in scenarios where behavioral and environmental factors heavily influence drivers' decision-making. Machine learning models play a crucial role in enhancing safety and informed decision-making by learning effective actions based on traffic conditions. However, training these models requires access to user data, which can compromise drivers' privacy and expose sensitive information. To address this issue, this study proposes a solution for generating synthetic driving condition data using a Federated Learning approach combined with Generative Adversarial Networks (GAN). This method enables model training across multiple federated learning clients, preserving data privacy by avoiding direct data sharing. By leveraging the Harmony dataset, similarity metrics such as Euclidean Distance and KL-Divergence were integrated into the GAN loss function to improve the quality of the generated synthetic data. The results demonstrate that the proposed approach successfully generates realistic driving condition data, supporting centralized model training while maintaining user privacy, showcasing its potential in privacy-conscious road safety applications.

Access Management for Content Delivery Networks: Measurements, Models, and Strategies

2025-04-09T17:11:05+00:00

We address the challenge of managing access to Content Delivery Networks (CDNs). In particular, we consider a scenario where users request tokens to access content, and one form of piracy consists in illegally sharing tokens. We focus on mitigating token misuse through performance analysis and statistical access pattern monitoring. Specifically, we examine how illegal token sharing impacts content delivery infrastructure and propose defining acceptable request limits to detect and block suspicious access patterns. Additionally, we introduce countermeasures against piracy, including selective quality degradation for users identified as engaging in illegal sharing, aiming to deter such behavior. Using queuing models, we quantify the impact of piracy on system performance across different scenarios. To validate our model, we perform statistical tests that compare real CDN traffic patterns with the expected request intervals in our proposed framework. These measures—defining access thresholds, quality degradation for unauthorized use, and statistical alignment checks—enhance CDN access management, preserving infrastructure integrity and the legitimate user experience while reducing operational costs.

Influence of Sequence Length and Geographic Representation on Optimal Prediction Architectures for Stolen Vehicle Geolocation

2025-03-12T01:50:45+00:00

When predicting the next geolocation of a stolen vehicle using external sensor data, such as speed radars, the challenge extends beyond the prediction itself to include determining the most suitable prediction architecture. While existing studies provide data that influence prediction performance, there is no consensus on the optimal architecture. Therefore, adopting a broader perspective to identify key criteria influencing the choice of architecture is essential. This study evaluates the shift in the optimal architecture depending on the length of the historical sequence and the format of geographic representation. The results reveal a shift in the optimal architecture, with the shift point being influenced by the type of geographic representation.

SISCMot: Situation Inference and Monitoring System for Intelligent Motorized Wheelchairs

2025-05-14T14:43:13+00:00

Assistive technologies aim to enhance independence and social inclusion for individuals with motor disabilities. Leveraging the Internet of Things (IoT), this study introduces SISCMot, an innovative monitoring and inference system tailored for intelligent motorized wheelchairs. SISCMot incorporates real-time data acquisition from the wheelchair’s electromechanical components, enabling context-aware analysis through IoT-connected dashboards. This system benefits a comprehensive range of stakeholders—including users, caregivers, technical support, and manufacturers—by providing critical insights into component performance, user mobility, and potential maintenance needs. Additionally, SISCMot offers a unique edge in preventive maintenance by monitoring key mechanical and electrical metrics, thereby extending the wheelchair’s operational lifespan. Evaluation based on the Technology Acceptance Model (TAM) confirms high usability and perceived value among target users, underscoring the system’s practical contribution to the assistive technology landscape.

Time-Weighted Correlation Approach to Identify High Delay Links in Internet Service Providers

2025-04-15T13:46:40+00:00

Companies and Internet Service Providers (ISPs) apply monitoring tools over network infrastructure, encompassing regular performance evaluations, with a primary focus on delivering crucial information about the current state of the network infrastructure and, consequently, the services running on it. However, these monitoring tools require ongoing development to handle more complex tasks, such as detecting performance issues. Within this context, this article proposes a mechanism for identifying high delays and communication links in the network that may cause these performance issues, using a temporally formulated Impact Score. This Score is based on data correlation techniques applied to information collected by monitoring tools. Experiments conducted with real data from the RNP Network indicate the efficiency of the proposal in identifying links impacting data communication, resulting in high end-to-end delays.

INTACT: Compact Storage of Data Streams in Mobile Devices to Unlock User Privacy at the Edge

2025-03-27T01:42:09+00:00

Data streams produced by mobile devices, such as smartphones, offer highly valuable sources of information to build ubiquitous services. Such data streams are generally uploaded and centralized to be processed by third parties, potentially exposing sensitive personal information. In this context, existing protection mechanisms, such as Location Privacy Protection Mechanisms (LPPMs), have been investigated. Alas, none of them have actually been implemented, nor deployed in real-life, in mobile devices to enforce user privacy at the edge. Moreover, the diversity of embedded sensors and the resulting data deluge makes it impractical to provision such services directly on mobiles, due to their constrained storage capacity, communication bandwidth and processing power. This article reports on the FLI technique, which leverages a piece-wise linear approximation technique to capture compact representations of data streams in mobile devices. Beyond the FLI storage layer, we introduce Divide & Stay, a new privacy preservation technique to execute Points of Interest (POIs) inference. Finally, we deploy both of them on Android and iOS as the INTACT framework, making a concrete step towards enforcing privacy and trust in ubiquitous computing systems.

Optimizing Compliance: Comparative Study of Data Laws and Privacy Frameworks

2025-04-10T18:33:09+00:00

Regarding privacy laws and digital globalization, understanding data regulation compliance and cross-jurisdictional challenges remains limited. To avoid administrative sanctions and protect user data, organizations and developers must bridge these gaps, navigating laws such as the General Data Protection Regulation (GDPR), the American Data Privacy and Protection Act (ADPPA), the General Data Protection Law (LGPD), and the Australian Privacy Act. This study focuses on creating a comprehensive compliance tool by investigating the similarities and nuances of these laws, as well as the challenges developers and organizations face in implementing Privacy by Design principles and ISO/IEC 29100 standards. Through a Systematic Literature Review (SLR) approach, topics of convergence and divergence among privacy laws and frameworks were pinpointed, as well as the challenges of implementing these laws in software. A survey was used to validate the challenges found in the SLR in the Brazilian context, in which most participants demonstrated a lack of knowledge regarding the LGPD. Lastly, we applied Framework Analysis to code and index key legislation points, allowing us to correlate them and develop a compliance-assistance tool. In the several contributions achieved, there is a deeper understanding of the privacy implications in a global context and its practical challenges, and also a practical guidance development, translating legal requirements into actions. Some limitations in this study lie in the interaction between selection and treatment in the survey, as participants' responses will not necessarily serve to generalize the challenges faced by all developers and organizations. In general, the contributions offer valuable theoretical and practical insights in the field of data privacy.

Synthetic Minority Over-sampling Technique for detecting Malicious Traffic targeting Internet of Things' devices

2025-04-18T12:40:56+00:00

This study proposes a multiclass machine learning approach for detecting 34 distinct types of cyberattacks in Internet of Things (IoT) traffic using the CICIoT2023 dataset. We evaluate the performance of lightweight classifiers—Bernoulli Naive Bayes, Decision Tree, Random Forest, and XGBoost—under highly imbalanced conditions. To address class imbalance and improve minority-class detection, we apply the Synthetic Minority Over-sampling Technique (SMOTE). In addition, we conduct hyperparameter tuning using RandomizedSearchCV and assess model performance using macro-average metrics, including recall, precision, and F1-score. Experimental results demonstrate that XGBoost and Random Forest, when optimized and combined with SMOTE, consistently achieves high and balanced detection rates across all classes. These findings suggest its applicability to real-world IoT intrusion detection scenarios, particularly in resource-constrained environments.

A Comprehensive Review of Techniques, Methods, Processes, Frameworks, and Tools for Privacy Requirements

2025-04-15T13:44:35+00:00

Context: Requirements Engineering (RE) relies on the collaboration of various roles—such as requirements engineers, stakeholders, and developers—and various techniques, methods, processes, frameworks, and tools. This makes RE a highly human-dependent process that benefits greatly from tool support. Understanding how these techniques, methods, processes, frameworks, and tools are applied across RE phases could provide valuable insights into ways to enhance the RE process, contributing to more successful outcomes. Objective: The primary objective of this study is to identify the techniques, methods, processes, frameworks, and tools applied across different requirements engineering phases—such as elicitation, analysis, specification, validation, and management—to address privacy requirements. Method: We conducted a systematic literature review (SLR) and identified 125 primary studies, and we also conducted a survey with 37 practitioners. Results: Our review identified a range of techniques, methods, processes, frameworks, and tools for addressing privacy requirements. Most studies were conducted in academic contexts, with the most frequently used tools being: PriS Method, Secure Tropos, LINDDUN, i* (i-star), STRAP (Structured Analysis for Privacy), Privacy by Design (PbD), and SQUARE. Additionally, over 75% of the studies applied these tools in the privacy requirements elicitation phase. In the industry, most of the techniques identified in the literature are not known or used by practitioners. Conclusion:This study provides a comprehensive analysis of techniques and tools for privacy requirements in RE, revealing a strong focus on academic contexts with limited industry application. Future research should explore the scalability and effectiveness of these tools in real-world environments, as well as the reasons why practitioners do not use them.

Data Privacy in Software Practice: Brazilian Developers’ Perspectives

2025-04-10T18:20:57+00:00

Data privacy is an essential principle of information security, aimed at protecting sensitive data from unauthorized access and information leaks. As software systems advance, the volume of personal information also grows exponentially. Therefore, incorporating privacy engineering practices during development is vital to ensure data integrity, confidentiality, and compliance with legal regulations, such as the General Data Protection Regulation (GDPR). However, there is a gap in understanding developers' awareness of data privacy, their perceptions of the implementation of privacy strategies, and the influence of organizational factors on this adoption. Thus, this paper aims to explore the level of awareness among Brazilian developers regarding data privacy and their perceptions of the implementation strategies adopted to ensure data privacy. Additionally, we seek to understand how organizational factors influence the adoption of data privacy practices. To this end, we surveyed 88 Brazilian developers with privacy-related work experience. We got 21 statements grouped into three topics to measure the Brazilian developers' awareness of data privacy in software. Our statistical analysis reveals substantial gaps between groups, e.g., developers have Direct v.s. Indirect data privacy-related work experience. We also reveal some data privacy strategies, e.g., Encryption, are both widely used and perceived as highly important, others, such as Turning off data collection, highlight strategies where ease of use does not necessarily lead to widespread adoption. Finally, we identified that the absence of dedicated privacy teams correlates with a lower perceived priority and less investment in tools. Even in organizations that recognize the importance of privacy. Our findings offer insights into how Brazilian developers perceive and implement data privacy practices, emphasizing the critical role organizational culture plays in decision-making regarding privacy. We hope that our findings will contribute to improving privacy practices within the software development community, particularly in contexts similar to Brazil.

Spectrum Defragmentation Window in SDM-EON Networks

2025-05-20T16:59:12+00:00

Space division multiplexing (SDM) technology expands the capacity of elastic optical networks (EONs) by adding spatial dimensions, positioning SDM-EONs as a strong candidate for future high-throughput infrastructures. However, SDM introduces new challenges, especially vertical fragmentation, where frequency slots become misaligned across multiple cores. This fragmentation decreases spectral efficiency, reduces resource availability, and increases connection blocking. This work proposes WDefrag, a novel RMSCA algorithm that tackles these issues through Slot Window Defragmentation, an original strategy developed in this study. WDefrag segments the spectrum into cost-evaluated windows and identifies regions where fragmentation most severely limits allocation. The algorithm reallocates resources locally, avoids unnecessary disruptions, and improves spectrum organization while managing crosstalk and fragmentation in both spatial and spectral dimensions. WDefrag operates in both proactive and reactive modes and adjusts window sizes to match traffic dynamics. Simulations compare it against non-defragmenting and state-of-the-art approaches. WDefrag outperforms these baselines by up to 30% in bandwidth blocking reduction, particularly in proactive scenarios. By applying cost-aware decisions and prioritizing fragmented regions that limit connectivity, WDefrag enhances spectrum utilization and delivers consistent performance improvements under real network demands.

Querying large video datasets: a systematic literature review

2025-09-01T14:54:06+00:00

Querying large-scale video datasets differs from querying short videos due to the inherent challenges in volume, velocity, and variety. In the last decade, this area has emerged thanks to the effectiveness of deep learning methods, new graphics processing units, new video databases, advances in distributed computing, among others. The main goal of querying video streams is to find the best balance between available hardware, software resources, and query latency, taking into account quality goals, constraints, and video configurations. Due to these challenges, many development methods, frameworks, and evaluation metrics have been proposed. As a result, this systematic literature review addresses a gap in the current body of knowledge. It covers ten years, from 2014 to 2024, and 4,248 papers, of which 99 were identified as relevant and used to answer the research questions on (i) processing methods, hardware architecture, and software, (ii) query languages, (iii) evaluation metrics, (iv) and available datasets. In addition, this review shows how this niche is promising and concerned with the rational use of available resources. Among the results, the following are highlighted: cheap detection models are very popular, smart IoT devices are very useful, distributed computing for video query applications is complex, system latency is essential, and there is no standard video query language. Current trends include the development of a standard video query language, in-memory computing, processing where data is produced, low-latency processing, and active learning for labeling objects. This original work shows a domain perspective, identifies problems and opportunities, and provides directions for future studies.

Disaster-FD: Federated Failure Detection in Disaster Scenarios

2025-05-02T15:30:59+00:00

This paper explores advanced features of Disaster-FD, a failure detector tailored for disaster-prone environments, with a specific focus on real-time monitoring of Internet of Things (IoT) networks. Leveraging federated monitoring, Disaster-FD enables the monitoring of geographically distributed regions to enhance network resilience. Inspired by Impact-FD, our proposed algorithm incorporates active monitoring and federated capabilities to ensure network reliability under adverse conditions. We conducted comprehensive experiments on the IoT-LAB platform to evaluate the robustness and resilience of Disaster-FD during potential disaster scenarios. These experiments assessed key parameters, including reliability thresholds, confidence levels, and impact factors, while ensuring efficient energy consumption and maintaining high network trust. Extensive evaluations, involving up to four geographically distinct regions in France and nearly a hundred IoT devices, demonstrate the effectiveness of Disaster-FD. Our findings highlight the potential of the algorithm to improve disaster response through enhanced IoT network monitoring, and we outline future directions for further development and optimization.

Composing State Machine Replication

2025-09-10T13:49:35+00:00

High availability is a fundamental requirement in large-scale distributed systems, where replication strategies are central in keeping applications operational despite a bounded number of failures. State Machine Replication (SMR) is one of the most widely adopted approaches for implementing highly available, fault-tolerant services, as it increases uptime while ensuring strong consistency. In recent years, research on SMR has yielded numerous variations tailored to enhance resilience, performance, and scalability. In this paper, we revisit SMR from a new perspective by introducing Composing State Machine Replication (CSMR), a method that enables fault-tolerant service composition. By composing SMRs, we promote the reuse of existing services to construct more complex and reliable systems. This modular approach fosters loosely coupled, flexible architectures, contributing to the theoretical foundations of SMR and aligning with common development practices in cloud computing and microservices. We formally define CSMR and demonstrate how composition can be used to extend existing SMR specifications with new features. For example, CSMR allows the semantics of a service operation to be extended by enabling different state machine replicas to execute complementary steps of the same operation. Additionally, SMR composition facilitates sharding and state partitioning by assigning disjoint state variables to separate SMRs. Beyond formalization, the paper provides illustrative examples of CSMR and introduces a high-level CSMR architecture that highlights the essential components, their responsibilities, and their interactions in supporting the composition process. To further demonstrate practicability, we present an API for building CSMR systems that combines RPC-based communication with declarative configuration in YAML format.

Generic Multicast: One Group Communication Primitive to Rule Them All

2025-09-10T13:46:19+00:00

Group communication primitives have a central role in modern computing infrastructures. They offer a panel of reliability and ordering guarantees for messages, enabling the implementation of complex distributed interactions. In particular, atomic broadcast is a pivotal abstraction for implementing fault-tolerant distributed services. This primitive allows disseminating messages across the system in a total order. There are two group communication primitives closely related to atomic broadcast. Atomic multicast permits targeting a subset of participants, possibly stricter than the whole system. Generic broadcast leverages the semantics of messages to order them only where necessary, that is, when they do not commute at the application level (a conflict). In this paper, we propose to combine all these primitives into a single, more general one, called generic multicast. We formally specify the guarantees offered by generic multicast and present efficient algorithms. Compared to prior works, our solutions offer appealing properties in terms of time and space complexity. In particular, when a run is conflict-free, that is no two messages conflict, a message is delivered after at most three message delays. We explain the logic of of our algorithms, detail their main invariants, and prove them correct. We also present a variation that delivers messages across the system in an order consistent with real-time at the cost of a message delay. This variation is particularly interesting to implement partially-replicated data storage systems

Adversary-Augmented Simulation for Fairness Evaluation and Defense in Hyperledger Fabric

2025-10-01T09:57:41+00:00

This paper presents an adversary model and a simulation framework specifically tailored for analyzing attacks on distributed systems composed of multiple distributed protocols, with a focus on assessing the security of blockchain networks. Our model classifies and constrains adversarial actions based on the assumptions of the target protocols—defined by failure models, communication models, and the fault tolerance thresholds of Byzantine Fault Tolerant (BFT) protocols. The goal is to study not only the intended effects of adversarial strategies but also their unintended side effects on critical system properties. We apply this framework to analyze fairness properties in a Hyperledger Fabric (HF) blockchain network. Our focus is on novel fairness attacks that involve coordinated adversarial actions across various HF services. Simulations show that even a constrained adversary can violate fairness with respect to specific clients (client fairness) and impact related guarantees (order fairness), which relate the reception order of transactions to their final order in the blockchain. This paper significantly extends our previous work by introducing and evaluating a mitigation mechanism specifically designed to counter transaction reordering attacks. We implement and integrate this defense into our simulation environment, demonstrating its effectiveness under diverse conditions.

Spreading Factor Allocation in LoRaWAN for Reliability and Delay-Constrained Smart Metering Applications

2025-07-29T17:41:10+00:00

Long Range Wide-Area Network (LoRaWAN) is a prominent IoT technology, and one of its operational parameters that significantly influences reliable transmission and packet delivery delay is Spreading Factor (SF). In this context, this paper develops a Spreading Factor Allocation scheme, named Delay and Reliability-aware Spreading Factor Allocation (DR-SFA), and evaluates it through simulations in comparison with four related solutions: Initial Spreading Factor Allocation (I-SFA), Adaptive Data Rate (ADR), I-SFA+ADR, and Collision-Aware Adaptive Data Rate (CA-ADR). The simulated scenarios model an Advanced Metering Infrastructure (AMI) executing Interval Meter Reading (IMR) and Power-Control Command (PCC) applications, with 200 to 1000 Smart Meters (SMs) distributed across an area of 56.25 km2. The results demonstrate that DR-SFA outperforms the alternative solutions by reducing the number of required Data Aggregation Points (DAPs) by up to 92.86%, while meeting the reliability and maximum delay requirements of the tested applications. Furthermore, DR-SFA successfully receives packets from meters located up to 2197.56 meters away in the scenario with 200 SMs, and achieves a Packet Delivery Ratio (PDR) of 99.49%, while also decreasing packet loss due to interference by 83.84% compared to ADR, and packet loss due to under sensitivity by 53.96% compared to I-SFA+ADR, in the scenario with 1000 SMs.

NES: Neural Embedding Squared

2024-11-28T01:17:14+00:00

In the fields of natural language processing (NLP) and machine learning, the quality and quantity of training data play a pivotal role in model performance. Textual data augmentation, a technique that artificially enhances the size of the training dataset by generating diverse yet semantically equivalent samples, has emerged as a crucial tool for overcoming data scarcity and improving the robustness of NLP models. However, the available solutions that achieve state-of-the-art performance require considerable computing power. This occurs because they use resource hungry machine learning models for each synthetic sentence generated. This paper introduces an approach to textual data augmentation, leveraging semantic representations to produce augmented data that not only expands the dataset but also understands the distribution of data points spatially. The approach requires less computing power by exploiting a fast prediction and spatial exploration in the embedding representation. In our experiments, it was able to double model performance while fixing class unbalance.

Enhancing Cloud Network Security with Innovative Time Series Analysis

2024-12-02T16:56:37+00:00

Cloud computing has revolutionized computing infrastructure abstraction and utilization, distinguished by its cost-effective and high-quality services. However, the challenge of securing cloud networks persists, mainly due to the broad exchange of data and the inherent complexity of these techniques. Anomaly detection emerges as a promising solution to improve cloud network safeness, presenting perception into system behavior and alerting operators for further actions. This paper offers a novel time series analysis method for detecting anomalies in cloud networks. Our technique employs innovative time series analysis techniques based on a matrix profile, and the Kneedle algorithm to identify multi-dimensional anomalous patterns within multiple features extracted from network traffic streams. To evaluate the efficacy of our approach, we implemented timestamp-based and index-based methods to two distinct datasets: the most widely used UNSW-NB15 and the recently introduced CICIoT2023 datasets. The results highlight the efficacy of our proposed method in identifying cloud network anomalies. It achieved an impressive accuracy of 99.6% and an F1-score of 99.8% using the timestamp-based analysis method. For the index-based analysis method, accuracy reached 98%, accompanied by an outstanding F1-score of 99.9%.

Automatic Inference of Brazilian Websites' Reliability for Combating Fake News: Domain and Geolocation Features

2025-01-30T17:14:54+00:00

Evaluating the reliability of websites that propagate news is critical in combating disinformation. Websites with low reliability often serve as the breeding ground for fake news that spreads rapidly across social networks. In response, this paper introduces an automatic evaluation approach to assessing the reliability of Brazilian websites by analyzing network-related features, eliminating the need for exhaustive content scanning.
Unlike previous methodologies focused on social network analysis, our approach leverages publicly available website features, including domain-related features, geolocation data, and TLS certificate attributes.
The paper proposes a supervised learning model and curates a comprehensive dataset comprising reliable and unreliable sites. Through rigorous training and evaluation using disjoint data, the model achieves an accuracy greater than 75%, effectively pinpointing reliable content websites.

Multi-Criteria, crosstalk-sensitive flexible topology approach for routing in SDM-EONs

2025-01-30T17:14:00+00:00

In recent years, the demand for high-capacity, flexible optical networks has driven significant advancements in network design, especially within Space-Division Multiplexing Elastic Optical Networks (SDM-EONs). These networks are expected to efficiently handle the increasing volume of data traffic while minimizing issues such as inter-core crosstalk and request blocking, which can degrade performance. To address these challenges, we propose an RMSSA (Routing, Modulation, Spectrum, and Space Assignment) approach specifically for SDM-EONs. Our method dynamically calculates link-weight across the network topology by considering a combination of critical network factors, enabling a flexible topology that self-balances network load prior to every incoming resource request. Next, our approach maps all slots in the spectrum guaranteed to not disrupt already allocated light-paths, while being under the crosstalk threshold for any desired modulation format, which is enabled by the adoption of precise slice-based crosstalk estimation (PS-XT). This multi-criteria, crosstalk-aware routing strategy significantly reduces inter-core crosstalk, leading to more efficient resource utilization and a considerable reduction in request blocking, ultimately enhancing the overall performance and reliability of SDM-EONs. Our comparative results show average reductions in request blocking of up to 77% for the load interval tested and up to 90% at lower loads, compared to approaches from the literature. Crosstalk was also reduced in over 36%, on average, and up to 60% at lower loads while fragmentation was mitigated in up to 11%.

From RockYou to RockYou2024: Analyzing Password Patterns Across Generations, Their Use in Industrial Systems and Vulnerability to Password Guessing Attacks

2024-10-07T18:30:10+00:00

Passwords are a common user authentication method, and must be safeguarded by effective security measures. However, there are many cases of compromised user credentials in data breaches. This work studies RockYou2024, a massive data breach that occurred in July 2024 and exposed over 9 billion passwords. We investigate the passwords with regard to their lengths, entropy, use of personal information and common strings, and evaluation from zxcvbn, as well as making a comparative assessment of the results with previous password databases, namely RockYou2021 and RockYou, which was leaked in 2009. This analysis found that the passwords from RockYou2021 and RockYou2024 are significantly more secure than those from RockYou, which suggests an improvement in password creation awareness and policies. It was also noted that RockYou2021 and RockYou2024 have similar statistical distributions in all the analyses conducted. We have also found that the country of origin for most passwords within these databases is most likely to be the United States of America. These datasets were searched for passwords that are often used in industrial systems, which pose potential security risks in critical infrastructure sectors. Finally, we also propose passBiRVAE, a contextualized Bidirectional Recurrent Neural Network , used to generate passwords based on the RockYou2024 database. Future works should make further improvements to the results obtained from this model. However, there is a risk of threats to the validity of these analyses.

Lightweight Malware Classification with FORTUNATE: Precision Meets Computational Efficiency

2025-02-13T14:15:49+00:00

After detecting a malicious artifact, classifying malware into specific families becomes an essential step to understand the threat's behavior, implement mitigation strategies, and develop proactive defenses. This task is particularly challenging due to the diversity of malware formats, the rapid evolution of obfuscation and packing techniques, as well as the scarcity of labeled data for training robust models. Additionally, the high volume of samples generated daily demands solutions that combine high accuracy and computational efficiency. Although transformer-based models are widely recognized as the state-of-the-art for sequence processing tasks, their high computational demands limit their practical application in resource-constrained environments. In this work, we present FORTUNATE, a lightweight framework that leverages LSTM networks with one-hot encoding to classify malware based on variable-length opcode sequences. The framework adopts an optimized opcode extraction process focused on reducing redundancies and representing data in compact vectors, minimizing computational costs. Experimental results indicate that FORTUNATE achieves accuracies of 99.82% for active malware and 99.81% for inactive malware, with an average classification time of only 56 ms per sample, significantly outperforming related works. The obtained results demonstrate that lightweight artificial intelligence approaches can deliver competitive performance in malware classification, especially in scenarios with computational constraints. FORTUNATE not only fills an important gap in malware classification but also establishes a foundation for future research aimed at optimizing the balance between accuracy, efficiency, and scalability.

Log parsers' performance on raw logs from Android devices

2025-02-10T13:24:14+00:00

Enhancing log file structure for improved analysis, commonly referred to as "Log Parsing'', holds significant importance in deciphering pertinent insights from software-generated records. This study undertakes a comprehensive comparison of ten parsing tools and models available within the Logpai collection, namely AEL, Brain, Drain, LFA, LogCluster, Logram, NuLog, SHISO, SLCT, and ULP focusing on raw logs sourced from Android Devices, extending a previous work. Our findings underscore a notable precision deficit in models lacking preprocessing steps, as existing tools encounter considerable challenges in managing untreated logs. Consequently, these tools exhibit suboptimal performance levels when analyzing information gleaned from raw Android Logs of the same origin as the reference logs. When analyzing other blocks, such as wifi networks, the difficulty of dealing with small variations in format was persistent.

Machine Learning-Based Strategy for Joint User Association and Resource Allocation in Next-Generation Networks

2025-03-03T12:09:33+00:00

This study presents an approach based on Reinforcement Learning (RL) to optimize the orchestration of User Association and Resource Allocation (UARA) mechanisms in next-generation heterogeneous networks, focusing on maximizing user satisfaction. The proposed strategy aims to improve the efficiency of these networks by overcoming operational challenges through user-centered adaptive algorithms. RL algorithms are utilized to rebalance the network load and optimize the distribution of radio resources among User Equipments (UEs), ultimately leading to improved service conditions. The results suggest that the strategic application of RL algorithms can lead to significant improvements compared to traditional methods, such as Max-SINR and Cell Range Expansion (CRE), reaching over 90% user satisfaction, highlighting the relevance of this research for next-generation networks.

Request Handling in Elastic Optical Data Center Networks: A Routing Algorithm Approach

2025-02-22T01:52:23+00:00

In recent years, worldwide communication has seen significant advancements driven by the growing demands of modern applications. These developments have introduced new requirements for data transfer, emphasizing the need for both high speeds and high efficiency. The cloud services model illustrates this shift, underscoring the importance of developing new mechanisms to handle increased data traffic. Beyond technological techniques, progress in the physical layer of networks is also crucial. It includes adopting spectrally and spatially flexible links that can precisely tune to the varying demands of network requests. In response to these challenges, this article proposes a routing algorithm tailored for Space-Division Multiplexing Data Center Elastic Optical Networks. The main objective of this algorithm is to improve network performance by maximizing the number of requests served efficiently. By taking advantage of the flexibility of optical networks, the proposed solution aims to enhance journal center capabilities while meeting the stringent requirements of modern communication systems. The potential of the proposed algorithm is seen in the Bandwidth Blocking Ratio, where it has better results than the other algorithms compared by up to two orders of magnitude, thus supporting the demands of the network.

Towards an efficient solution to mitigate the forest fire problem based on unmanned aerial vehicles and wireless sensors

2025-02-11T21:10:58+00:00

One of the leading global challenges that the society faces worldwide is related to the forest fire problem, which generates financial losses, threatens ecological systems, and affects public security, putting human and animal life at risk. Despite recent efforts to mitigate the forest fire problem, providing a higher accuracy rate for detecting fire, with a quick response time, without impacting the alert process is still a challenging R&D question that must be investigated. To advance this research front, we propose a solution to detect and monitor forest fires, called DF-Fire, using a UAV (Unmanned Aerial Vehicle) and WSN (Wireless Sensor Network). For this, a deep learning architecture is modeled to carry out the fire detection process. In addition, to cover the area of interest, DF-Fire has a flight plan based on the information that the WSN disseminates. DF-Fire has been evaluated on real devices to prove its efficiency and, when compared to other benchmarking solutions, our solution has advanced in state of the art by: (i) increasing the hit rate to detect the fire; (ii) reduce the response time; and (iii) reduce overhead in processing time without impacting the alert process. Also, DF-Fire takes advantage of the sensors’ information to provide efficiency in the flight plan and correlate them to monitor how the fire spreads.

A Comparative Evaluation of Symmetric Cryptography Algorithms for Resource-Constrained Devices

2025-03-03T12:09:53+00:00

Data security in the Internet of Things (IoT) is crucial for protecting both the devices and the data they transmit over the network. Nevertheless, security is often overlooked in this context, leaving systems vulnerable to cyberattacks that can compromise information confidentiality and integrity. This work focuses on a use case of an environment remote monitoring system for disaster prevention, in which information must be confidential and intact. The AES (Advanced Encryption Standard) and Speck cryptographic algorithm families are evaluated in both traditional and memory-optimized implementations, targeting data confidentiality. The algorithms are assessed through practical experiments on two resource-constrained hardware platforms. Transmission throughput, estimated energy consumption, execution time, and memory usage are evaluated. Results show that the Speck family executes more quickly, has lower estimated energy consumption, and occupies less memory space than AES in both platforms.

ONSPRIDE: An Ontology-based Framework for Privacy by Design in Distributed Networks Apps

2025-02-10T13:22:50+00:00

With the advancement of technologies for data registration in distributed networks, the concern of users and developers of computerized solutions with the privacy of sensitive data has increased. Thus, this work addresses a conceptual solution for an ontology-based framework so that any entity willing to provide a service using Distributed Ledger Technology (DLT) networks can model the set of privacy attributes of its system according to the business rules of its service. The solution proposed in this work encompasses the development of an architecture aimed at providing computational support for the privacy design of the actors involved in the offering and consumption of services implemented in DLTs. The architecture also includes a framework called ONSPRIDE, which uses previously stored domain ontologies to translate business rules into requirements and privacy. We conducted a proof of context by comparing the performance of two Hyperledger Fabric networks. For this purpose, we conducted a controlled experiment in which both networks operate a smart contract that manages attendance records for outdoor events. The main difference between the networks is that one uses a Certificate Authority (CA) to issue access certificates, while the other issues certificates manually. We compared the results obtained through the reports generated by the Hyperledger Caliper tool. In addition, the performance of the initialization and connection of agents in a Self-Sovereign Identity system was measured. The results of this study provide valuable insight that can help developers choose the most suitable ledger type for their Hyperledger projects and support decision-making regarding adopting a Self-Sovereign Identity system.

FORENSICS: Deciphering and Detecting Malware Through Variable-Length Instruction Sequences

2025-06-26T18:51:12+00:00

The increasing complexity of contemporary malware, driven by the use of advanced evasion and obfuscation techniques, combined with the rapid growth in the number of new variants emerging continuously, undermines the effectiveness of traditional signature-based detection mechanisms. In response to this scenario, this work proposes FORENSICS (Framework fOr malwaRe dEtectioN baSed on InstruCtion Sequences), an innovative deep learning-based framework that employs variable-length instruction sequences to detect malware efficiently and accurately. By integrating Natural Language Processing (NLP) techniques with Long Short-Term Memory (LSTM) neural networks, FORENSICS analyzes opcode sequences extracted from real-world malware and benign software artifacts. The framework introduces optimized methods for opcode extraction and representation, significantly reducing computational overhead while preserving detection performance. FORENSICS achieved 99.91% accuracy, 99.99% precision, and detection times ranging from 8 to 17 milliseconds, outperforming several state-of-the-art approaches across multiple metrics. Additionally, the framework demonstrated robustness in identifying zero-day malware samples, confirming its effectiveness in real-world cybersecurity scenarios. A new balanced dataset comprising over 40,000 labeled samples was created and made publicly available, facilitating reproducibility and encouraging further research. These results position FORENSICS as a robust, scalable, and highly effective solution for malware detection in modern threat landscapes.

Modeling Interest Networks in Urban Areas: A Comparative Study of Google Places and Foursquare Across Countries

2025-02-23T21:41:00+00:00

Location-Based Social Networks (LBSNs) are valuable for understanding urban behavior and providing useful data on user preferences. Modeling their data into graphs like interest networks (iNETs) offers important insights for urban area recommendations, mobility forecasting, and public policy development. This study uses check-ins and venue reviews to compare the iNETs resulting from two distinct LBSNs, Foursquare and Google Places. Although these two LBSNs differ in nature, with data varying in regularity and purpose, their resulting iNETs reveal similar urban behavior patterns. When analyzing the impact of socioeconomic, political, and geographic factors on iNET edges — each edge representing users' interests in a pair of regions — only geographic factors showed a significant influence. When studying the granularity of area sizes to model iNETs, we highlight important trade-offs between larger and smaller sizes. Additionally, we propose a methodology to identify clusters of geographically neighboring areas where user interest is strongest, which can be advantageous for understanding urban space usage.