arXiv subjects

Eva Maia

Publications and source records attributed to Eva Maia.

At least 19 recordsLinked to original sources

arXiv2026

Machine Unlearning for the XGBoost Model with Network Intrusion Datasets

Machine Unlearning (MU) has emerged as an important technique for removing specific data points from trained models without requiring full retraining. However, most existing MU research focuses on deep learning and image data, leaving a gap in the domain of network intrusion detection, which relies heavily on tabular data. This work introduces XGBoost-Forget, an unlearning approach for the XGBoost model, to address this gap. The approach is evaluated on two tabular Network Intrusion (NI) datasets, IoT-23 and GeNIS, using multiple metrics to assess model performance, unlearning efficiency, and forgetting quality. The results show that XGBoost-Forget maintains predictive performance close to the original model while providing significantly faster unlearning, demonstrating its potential for MU in tabular NI settings.

cs.LG↗

arXiv2026

Influence Factors on RAG Poisoning

Retrieval-Augmented Generation (RAG) systems enhance large language models by grounding responses in retrieved documents from external knowledge sources at inference time. However, this reliance on retrieved content introduces vulnerabilities to poisoning attacks, in which adversarial documents can manipulate both the retrieval process and the generated outputs. This paper investigates poisoning robustness in RAG through a full factorial experimental study covering 432 configurations. We analyze the impacts of dataset, retriever type, retrieval depth, database composition, chunking strategy, and generator model on retrieval-level and generation-level metrics. The results show that retriever architecture, dataset, and retrieval depth are the strongest factors affecting poisoning exposure, while generator choice and database composition have a major impact on downstream attack success. Dense and graph-based retrievers generally improve robustness relative to BM25, whereas larger retrieval depth increases the likelihood of retrieving poisoned passages. We further show that replicating poisoned content across multiple databases amplifies adversarial influence, while additional clean sources can mitigate it. These findings highlight that poisoning vulnerability in RAG is not attributable to a single component, but instead arises from the interaction of retrieval, generation, and knowledge-base configuration.

cs.CR↗

arXiv2026

Evaluating Local Explainability Metrics for Machine Learning Models on Tabular Data

Despite the wide use of explainability techniques to attempt to understand the behavior of Artificial Intelligence (AI), the generated explanations may not always be reliable. An explanation can appear plausible to humans but fail to capture the internal reasoning of a model, particularly when dealing with complex tabular data. This paper studies the trustworthiness of local explainability techniques when applied to complex tabular classification tasks, considering evaluated metrics for three main properties: faithfulness to the model's predictions, robustness to input data variations, and complexity of the explanation itself. A benchmark was performed for Local Interpretable Model-Agnostic Explanations (LIME), Kernel SHapley Additive exPlanations (SHAP), and Feature Ablation techniques, across 32 datasets and different types of machine learning models. Model performance ranges were analyzed to identify two groups: consensus-correct, which are samples that all models predicted correctly, and consensus-wrong, samples that all models predicted incorrectly. The obtained results demonstrate that that the explanations are not always correlated with a model's predictive performance. Instead, dataset complexity and feature distributions seem to be the main factors affecting explanation quality and reliability.

cs.LG↗

arXiv2026

Towards Privacy-Preserving Federated Learning using Hybrid Homomorphic Encryption

Federated Learning (FL) enables collaborative training while keeping sensitive data on clients' devices, but local model updates can still leak private information. Hybrid Homomorphic Encryption (HHE) has recently been applied to FL to mitigate client overhead while preserving privacy. However, existing HHE-FL systems rely on a single homomorphic key pair shared across all clients, which forces them to assume an unrealistically weak threat model: if a client misbehaves or intercepts another's traffic, private updates can be exposed. We eliminate this weakness by integrating two alternative key protection mechanisms into the HHE-FL workflow. The first is masking, where client keys are blinded before homomorphic encryption and later unblinded homomorphically by the server. The second is RSA encapsulation, where homomorphically encrypted keys are additionally wrapped under the server's RSA public key. These countermeasures prevent key misuse by other clients and extend HHE-FL security to adversarial settings with malicious participants. We implement both approaches on top of the Flower framework using the PASTA/BFV HHE scheme and evaluate them on the MNIST dataset with 12 clients. Results show that both mechanisms preserve model accuracy while adding minimal overhead: masking incurs negligible cost, and RSA encapsulation introduces only modest runtime and communication overhead.

cs.CR↗

arXiv2026

Machine Learning Transferability for Malware Detection

Malware continues to be a predominant operational risk for organizations, especially when obfuscation techniques are used to evade detection. Despite the ongoing efforts in the development of Machine Learning (ML) detection approaches, there is still a lack of feature compatibility in public datasets. This limits generalization when facing distribution shifts, as well as transferability to different datasets. This study evaluates the suitability of different data preprocessing approaches for the detection of Portable Executable (PE) files with ML models. The preprocessing pipeline unifies EMBERv2 (2,381-dim) features datasets, trains paired models under two training setups: EMBER + BODMAS and EMBER + BODMAS + ERMDS. Regarding model evaluation, both EMBER + BODMAS and EMBER + BODMAS + ERMDS models are tested against TRITIUM, INFERNO and SOREL-20M. ERMDS is also used for testing for the EMBER + BODMAS setup.

cs.CR↗

arXiv2025

Revisiting Network Traffic Analysis: Compatible network flows for ML models

To ensure that Machine Learning (ML) models can perform a robust detection and classification of cyberattacks, it is essential to train them with high-quality datasets with relevant features. However, it can be difficult to accurately represent the complex traffic patterns of an attack, especially in Internet-of-Things (IoT) networks. This paper studies the impact that seemingly similar features created by different network traffic flow exporters can have on the generalization and robustness of ML models. In addition to the original CSV files of the Bot-IoT, IoT-23, and CICIoT23 datasets, the raw network packets of their PCAP files were analysed with the HERA tool, generating new labelled flows and extracting consistent features for new CSV versions. To assess the usefulness of these new flows for intrusion detection, they were compared with the original versions and were used to fine-tune multiple models. Overall, the results indicate that directly analysing and preprocessing PCAP files, instead of just using the commonly available CSV files, enables the computation of more relevant features to train bagging and gradient boosting decision tree ensembles. It is important to continue improving feature extraction and feature selection processes to make different datasets more compatible and enable a trustworthy evaluation and comparison of the ML models used in cybersecurity solutions.

cs.CR↗

arXiv2025

Binary and Multiclass Cyberattack Classification on GeNIS Dataset

The integration of Artificial Intelligence (AI) in Network Intrusion Detection Systems (NIDS) is a promising approach to tackle the increasing sophistication of cyberattacks. However, since Machine Learning (ML) and Deep Learning (DL) models rely heavily on the quality of their training data, the lack of diverse and up-to-date datasets hinders their generalization capability to detect malicious activity in previously unseen network traffic. This study presents an experimental validation of the reliability of the GeNIS dataset for AI-based NIDS, to serve as a baseline for future benchmarks. Five feature selection methods, Information Gain, Chi-Squared Test, Recursive Feature Elimination, Mean Absolute Deviation, and Dispersion Ratio, were combined to identify the most relevant features of GeNIS and reduce its dimensionality, enabling a more computationally efficient detection. Three decision tree ensembles and two deep neural networks were trained for both binary and multiclass classification tasks. All models reached high accuracy and F1-scores, and the ML ensembles achieved slightly better generalization while remaining more efficient than DL models. Overall, the obtained results indicate that the GeNIS dataset supports intelligent intrusion detection and cyberattack classification with time-based and quantity-based behavioral features.

cs.CR↗

arXiv2025

Adversarially Robust and Interpretable Magecart Malware Detection

Magecart skimming attacks have emerged as a significant threat to client-side security and user trust in online payment systems. This paper addresses the challenge of achieving robust and explainable detection of Magecart attacks through a comparative study of various Machine Learning (ML) models with a real-world dataset. Tree-based, linear, and kernel-based models were applied, further enhanced through hyperparameter tuning and feature selection, to distinguish between benign and malicious scripts. Such models are supported by a Behavior Deterministic Finite Automaton (DFA) which captures structural behavior patterns in scripts, helping to analyze and classify client-side script execution logs. To ensure robustness against adversarial evasion attacks, the ML models were adversarially trained and evaluated using attacks from the Adversarial Robustness Toolbox and the Adaptative Perturbation Pattern Method. In addition, concise explanations of ML model decisions are provided, supporting transparency and user trust. Experimental validation demonstrated high detection performance and interpretable reasoning, demonstrating that traditional ML models can be effective in real-world web security contexts.

cs.CR↗

arXiv2025

SPATA: Systematic Pattern Analysis for Detailed and Transparent Data Cards

Due to the susceptibility of Artificial Intelligence (AI) to data perturbations and adversarial examples, it is crucial to perform a thorough robustness evaluation before any Machine Learning (ML) model is deployed. However, examining a model's decision boundaries and identifying potential vulnerabilities typically requires access to the training and testing datasets, which may pose risks to data privacy and confidentiality. To improve transparency in organizations that handle confidential data or manage critical infrastructure, it is essential to allow external verification and validation of AI without the disclosure of private datasets. This paper presents Systematic Pattern Analysis (SPATA), a deterministic method that converts any tabular dataset to a domain-independent representation of its statistical patterns, to provide more detailed and transparent data cards. SPATA computes the projection of each data instance into a discrete space where they can be analyzed and compared, without risking data leakage. These projected datasets can be reliably used for the evaluation of how different features affect ML model robustness and for the generation of interpretable explanations of their behavior, contributing to more trustworthy AI.

cs.LG↗

arXiv2025

Federated Learning: An approach with Hybrid Homomorphic Encryption

Federated Learning (FL) is a distributed machine learning approach that promises privacy by keeping the data on the device. However, gradient reconstruction and membership-inference attacks show that model updates still leak information. Fully Homomorphic Encryption (FHE) can address those privacy concerns but it suffers from ciphertext expansion and requires prohibitive overhead on resource-constrained devices. We propose the first Hybrid Homomorphic Encryption (HHE) framework for FL that pairs the PASTA symmetric cipher with the BFV FHE scheme. Clients encrypt local model updates with PASTA and send both the lightweight ciphertexts and the PASTA key (itself BFV-encrypted) to the server, which performs a homomorphic evaluation of the decryption circuit of PASTA and aggregates the resulting BFV ciphertexts. A prototype implementation, developed on top of the Flower FL framework, shows that on independently and identically distributed MNIST dataset with 12 clients and 10 training rounds, the proposed HHE system achieves 97.6% accuracy, just 1.3% below plaintext, while reducing client upload bandwidth by over 2,000x and cutting client runtime by 30% compared to a system based solely on the BFV FHE scheme. However, server computational cost increases by roughly 15621x for each client participating in the training phase, a challenge to be addressed in future work.

cs.CR↗

arXiv2025

MeAJOR Corpus: A Multi-Source Dataset for Phishing Email Detection

Phishing emails continue to pose a significant threat to cybersecurity by exploiting human vulnerabilities through deceptive content and malicious payloads. While Machine Learning (ML) models are effective at detecting phishing threats, their performance largely relies on the quality and diversity of the training data. This paper presents MeAJOR (Merged email Assets from Joint Open-source Repositories) Corpus, a novel, multi-source phishing email dataset designed to overcome critical limitations in existing resources. It integrates 135894 samples representing a broad number of phishing tactics and legitimate emails, with a wide spectrum of engineered features. We evaluated the dataset's utility for phishing detection research through systematic experiments with four classification models (RF, XGB, MLP, and CNN) across multiple feature configurations. Results highlight the dataset's effectiveness, achieving 98.34% F1 with XGB. By integrating broad features from multiple categories, our dataset provides a reusable and consistent resource, while addressing common challenges like class imbalance, generalisability and reproducibility.

cs.CR↗

arXiv2025

Enhancing JavaScript Malware Detection through Weighted Behavioral DFAs

This work addresses JavaScript malware detection to enhance client-side web application security with a behavior-based system. The ability to detect malicious JavaScript execution sequences is a critical problem in modern web security as attack techniques become more sophisticated. This study introduces a new system for detecting JavaScript malware using a Deterministic Finite Automaton (DFA) along with a weighted-behavior system, which we call behavior DFA. This system captures malicious patterns and provides a dynamic mechanism to classify new sequences that exhibit partial similarity to known attacks, differentiating them between benign, partially malicious, and fully malicious behaviors. Experimental evaluation on a dataset of 1,058 sequences captured in a real-world environment demonstrates the capability of the system to detect and classify threats effectively, with the behavior DFA successfully identifying exact matches and partial similarities to known malicious behaviors. The results highlight the adaptability of the system in detecting emerging threats while maintaining transparency in decision making.

cs.CR↗

arXiv2025

Enhancing Large Language Models with Faster Code Preprocessing for Vulnerability Detection

The application of Artificial Intelligence has become a powerful approach to detecting software vulnerabilities. However, effective vulnerability detection relies on accurately capturing the semantic structure of code and its contextual relationships. Given that the same functionality can be implemented in various forms, a preprocessing tool that standardizes code representation is important. This tool must be efficient, adaptable across programming languages, and capable of supporting new transformations. To address this challenge, we build on the existing SCoPE framework and introduce SCoPE2, an enhanced version with improved performance. We compare both versions in terms of processing time and memory usage and evaluate their impact on a Large Language Model (LLM) for vulnerability detection. Our results show a 97.3\% reduction in processing time with SCoPE2, along with an improved F1-score for the LLM, solely due to the refined preprocessing approach.

cs.SE↗

arXiv2025

Evaluating LLaMA 3.2 for Software Vulnerability Detection

Deep Learning (DL) has emerged as a powerful tool for vulnerability detection, often outperforming traditional solutions. However, developing effective DL models requires large amounts of real-world data, which can be difficult to obtain in sufficient quantities. To address this challenge, DiverseVul dataset has been curated as the largest dataset of vulnerable and non-vulnerable C/C++ functions extracted exclusively from real-world projects. Its goal is to provide high-quality, large-scale samples for training DL models. However, during our study several inconsistencies were identified in the raw dataset while applying pre-processing techniques, highlighting the need for a refined version. In this work, we present a refined version of DiverseVul dataset, which is used to fine-tune a large language model, LLaMA 3.2, for vulnerability detection. Experimental results show that the use of pre-processing techniques led to an improvement in performance, with the model achieving an F1-Score of 66%, a competitive result when compared to our baseline, which achieved a 47% F1-Score in software vulnerability detection.

cs.LG↗

arXiv2025

A Novel Approach to Network Traffic Analysis: the HERA tool

Cybersecurity threats highlight the need for robust network intrusion detection systems to identify malicious behaviour. These systems rely heavily on large datasets to train machine learning models capable of detecting patterns and predicting threats. In the past two decades, researchers have produced a multitude of datasets, however, some widely utilised recent datasets generated with CICFlowMeter contain inaccuracies. These result in flow generation and feature extraction inconsistencies, leading to skewed results and reduced system effectiveness. Other tools in this context lack ease of use, customizable feature sets, and flow labelling options. In this work, we introduce HERA, a new open-source tool that generates flow files and labelled or unlabelled datasets with user-defined features. Validated and tested with the UNSW-NB15 dataset, HERA demonstrated accurate flow and label generation.

cs.CR↗

arXiv2024

Flow Exporter Impact on Intelligent Intrusion Detection Systems

High-quality datasets are critical for training machine learning models, as inconsistencies in feature generation can hinder the accuracy and reliability of threat detection. For this reason, ensuring the quality of the data in network intrusion detection datasets is important. A key component of this is using reliable tools to generate the flows and features present in the datasets. This paper investigates the impact of flow exporters on the performance and reliability of machine learning models for intrusion detection. Using HERA, a tool designed to export flows and extract features, the raw network packets of two widely used datasets, UNSW-NB15 and CIC-IDS2017, were processed from PCAP files to generate new versions of these datasets. These were compared to the original ones in terms of their influence on the performance of several models, including Random Forest, XGBoost, LightGBM, and Explainable Boosting Machine. The results obtained were significant. Models trained on the HERA version of the datasets consistently outperformed those trained on the original dataset, showing improvements in accuracy and indicating a better generalisation. This highlighted the importance of flow generation in the model's ability to differentiate between benign and malicious traffic.

cs.CR↗

arXiv2024

Network Simulation with Complex Cyber-attack Scenarios

Network Intrusion Detection (NID) systems can benefit from Machine Learning (ML) models to detect complex cyber-attacks. However, to train them with a great amount of high-quality data, it is necessary to perform reliable simulations of multiple interacting machines. This paper presents a network simulation solution for the creation of NID datasets with complex attack scenarios. This solution was integrated in the Airbus CyberRange platform to benefit from its simulation capabilities of generating benign and malicious traffic patterns that represent realistic cyber-attacks targeting a computer network. A realistic vulnerable network topology was configured in the CyberRange and three different attack scenarios were implemented: Man-in-the-Middle (MitM), Denial-of-Service (DoS), and Brute-Force (BF).

cs.CR↗

arXiv2024

Intelligent Green Efficiency for Intrusion Detection

Artificial Intelligence (AI) has emerged in popularity recently, recording great progress in various industries. However, the environmental impact of AI is a growing concern, in terms of the energy consumption and carbon footprint of Machine Learning (ML) and Deep Learning (DL) models, making essential investigate Green AI, an attempt to reduce the climate impact of AI systems. This paper presents an assessment of different programming languages and Feature Selection (FS) methods to improve computation performance of AI focusing on Network Intrusion Detection (NID) and cyber-attack classification tasks. Experiments were conducted using five ML models - Random Forest, XGBoost, LightGBM, Multi-Layer Perceptron, and Long Short-Term Memory - implemented in four programming languages - Python, Java, R, and Rust - along with three FS methods - Information Gain, Recursive Feature Elimination, and Chi-Square. The obtained results demonstrated that FS plays an important role enhancing the computational efficiency of AI models without compromising detection accuracy, highlighting languages like Python and R, that benefit from a rich AI libraries environment. These conclusions can be useful to design efficient and sustainable AI systems that still provide a good generalization and a reliable detection.

cs.CR↗