Mehdi Zakroum
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Mehdi Zakroum has a total of 29 co-authors. Here are the top co-authors by number of shared publications:
- Mounir Ghogho
Mohammed VI Polytechnic University
8 shared publications Isabelle Chrisment
Universit de Lorraine, Nancy, France
2 shared publications- Jerome Francois
University of Luxembourg, Luxembourg, Luxembourg
2 shared publications - Abdelghani Ghanem
TICLab, College of Engineering & Architecture, International University of Rabat, Rabat 11103, Morocco
1 shared publication Abdelkader Lahmadi
Inria – LORIA – University of Lorraine
1 shared publication- Abdelkader Lahmadi
Université de Lorraine, CNRS, Inria, LORIA, Nancy, France
1 shared publication
Latest Publications
- 7 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Machine Learning for the Automation of Cyber-threat Monitoring and Inference
Apprentissage automatique pour l'automatisation de la surveillance et l'inférence des cybermenaces: Au cours des dernières décennies, les cyber-menaces ont connu une augmentation significative et continuent de croître de façon exponentielle. Les opérateurs de réseau et les praticiens de la sécurité s'efforcent constamment d'automatiser leurs stratégies de défense contre les cyberincidents à grande échelle et les événements particuliers à plus petite échelle ciblant leurs réseaux. Améliorer la surveillance des événements de sécurité et détecter les attaques à un stade précoce sont des éléments clés pour prévenir des éventuels dommages ou au moins atténuer leurs impacts. Le trafic enregistré par les capteurs réseau tels que les télescopes réseau, également connus sous le nom de darknets, constitue une riche source de renseignements sur la cybersécurité. Les données enregistrées par ces capteurs incluent différents types de trafic allant du trafic bénin comme les analyses régulières effectuées par les organisations à des fins statistiques, aux cyber-incidents malveillants tels que la propagation de vers, les analyses de vulnérabilité et les paquets de rétrodiffusion en relation avec les attaques par déni de service. Ces données pourraient être exploitées pour automatiser et améliorer les solutions de surveillance des cyber-menaces ainsi que pour modéliser et prédire les attaques. Pour cela, cette thèse combine des travaux de recherche sur les sujets saillants de la surveillance des cyber-menaces et de la classification et de la prévision des cyber-attaques.
- 7 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Machine Learning for the Automation of Cyber-threat Monitoring and Inference
Apprentissage automatique pour l'automatisation de la surveillance et l'inférence des cybermenaces: Au cours des dernières décennies, les cyber-menaces ont connu une augmentation significative et continuent de croître de façon exponentielle. Les opérateurs de réseau et les praticiens de la sécurité s'efforcent constamment d'automatiser leurs stratégies de défense contre les cyberincidents à grande échelle et les événements particuliers à plus petite échelle ciblant leurs réseaux. Améliorer la surveillance des événements de sécurité et détecter les attaques à un stade précoce sont des éléments clés pour prévenir des éventuels dommages ou au moins atténuer leurs impacts. Le trafic enregistré par les capteurs réseau tels que les télescopes réseau, également connus sous le nom de darknets, constitue une riche source de renseignements sur la cybersécurité. Les données enregistrées par ces capteurs incluent différents types de trafic allant du trafic bénin comme les analyses régulières effectuées par les organisations à des fins statistiques, aux cyber-incidents malveillants tels que la propagation de vers, les analyses de vulnérabilité et les paquets de rétrodiffusion en relation avec les attaques par déni de service. Ces données pourraient être exploitées pour automatiser et améliorer les solutions de surveillance des cyber-menaces ainsi que pour modéliser et prédire les attaques. Pour cela, cette thèse combine des travaux de recherche sur les sujets saillants de la surveillance des cyber-menaces et de la classification et de la prévision des cyber-attaques.
- Open Access
- 13 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 194-205
LLM-CVX: A Benchmarking Framework for Assessing the Offensive Potential of LLMs in Exploiting CVEs
- 13 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 194-205
Open AccessLLM-CVX: A Benchmarking Framework for Assessing the Offensive Potential of LLMs in Exploiting CVEs
- Open Access
Cross-Lingual Multi-Granularity Framework for Interpretable Parkinson's Disease Diagnosis from Speech
Parkinson's Disease (PD) affects over 10 million people worldwide, with speech impairments in up to 89% of patients. Current speech-based detection systems analyze entire utterances, potentially overlooking the diagnostic value of specific phonetic elements. We developed a granularity-aware approach for multilingual PD detection using an automated pipeline that extracts time-aligned phonemes, syllables, and words from recordings. Using Italian, Spanish, and English datasets, we implemented a bidirectional LSTM with multi-head attention to compare diagnostic performance across the different granularity levels. Phoneme-level analysis achieved superior performance with AUROC of 93.78% +- 2.34% and accuracy of 92.17% +- 2.43%. This demonstrates enhanced diagnostic capability for cross-linguistic PD detection. Importantly, attention analysis revealed that the most informative speech features align with those used in established clinical protocols: sustained vowels (/a/, /e/, /o/, /i/) at phoneme level, diadochokinetic syllables (/ta/, /pa/, /la/, /ka/) at syllable level, and /pataka/ sequences at word level. Source code will be available at https://github.com/jetliqs/clearpd.
Open AccessCross-Lingual Multi-Granularity Framework for Interpretable Parkinson's Disease Diagnosis from Speech
Parkinson's Disease (PD) affects over 10 million people worldwide, with speech impairments in up to 89% of patients. Current speech-based detection systems analyze entire utterances, potentially overlooking the diagnostic value of specific phonetic elements. We developed a granularity-aware approach for multilingual PD detection using an automated pipeline that extracts time-aligned phonemes, syllables, and words from recordings. Using Italian, Spanish, and English datasets, we implemented a bidirectional LSTM with multi-head attention to compare diagnostic performance across the different granularity levels. Phoneme-level analysis achieved superior performance with AUROC of 93.78% +- 2.34% and accuracy of 92.17% +- 2.43%. This demonstrates enhanced diagnostic capability for cross-linguistic PD detection. Importantly, attention analysis revealed that the most informative speech features align with those used in established clinical protocols: sustained vowels (/a/, /e/, /o/, /i/) at phoneme level, diadochokinetic syllables (/ta/, /pa/, /la/, /ka/) at syllable level, and /pataka/ sequences at word level. Source code will be available at https://github.com/jetliqs/clearpd.
- Open Access
- 2 June 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 75191-75214
Spatio-Temporal Forecasting: A Survey of Data-Driven Models Using Exogenous Data
Forecasting Spatio-Temporal processes has been attracting a great deal of interest within the research community. In this context, there is an increasing trend of proposing and improving methodologies to gather and use vast amounts of Spatio-Temporal data. Spatio Temporal Forecasting (STF) problems present complex interactions and non-linearities as the temporal dimension and the spatial one are usually entangled. To address these problems, statistical, Machine Learning based, and Deep Learning based models are introduced and developed. The use of exogenous data has proven to be beneficial in many STF models. Various techniques of incorporating exogenous data in STF problems have been proposed in the literature. This survey aims at providing a systematic review of the data-driven STF models, with a focus on those that incorporate exogenous data. We first investigate the data properties, including their dynamics, types and representations. Next, we propose a new taxonomy of the reviewed models and inspect the different complexities of STF problems. Exogenous data incorporation techniques are then presented and analyzed. We conclude our paper by highlighting the current open challenges and future research directions.
- 2 June 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 75191-75214
Open AccessSpatio-Temporal Forecasting: A Survey of Data-Driven Models Using Exogenous Data
Forecasting Spatio-Temporal processes has been attracting a great deal of interest within the research community. In this context, there is an increasing trend of proposing and improving methodologies to gather and use vast amounts of Spatio-Temporal data. Spatio Temporal Forecasting (STF) problems present complex interactions and non-linearities as the temporal dimension and the spatial one are usually entangled. To address these problems, statistical, Machine Learning based, and Deep Learning based models are introduced and developed. The use of exogenous data has proven to be beneficial in many STF models. Various techniques of incorporating exogenous data in STF problems have been proposed in the literature. This survey aims at providing a systematic review of the data-driven STF models, with a focus on those that incorporate exogenous data. We first investigate the data properties, including their dynamics, types and representations. Next, we propose a new taxonomy of the reviewed models and inspect the different complexities of STF problems. Exogenous data incorporation techniques are then presented and analyzed. We conclude our paper by highlighting the current open challenges and future research directions.
- Open Access
- 29 March 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 90749-90765
Self-Supervised Latent Representations of Network Flows and Application to Darknet Traffic Classification
Characterizing network flows is essential for security operators to enhance their awareness about cyber-threats targeting their networks. The automation of network flow characterization with machine learning has received much attention in recent years. To this aim, raw network flows need to be transformed into structured and exploitable data. In this research work, we propose a method to encode raw network flows into robust latent representations exploitable in downstream tasks. First, raw network flows are transformed into graph-structured objects capturing their topological aspects (packet-wise transitional patterns) and features (used protocols, packets’ flags, etc.). Then, using self-supervised learning techniques like Graph Auto-Encoders and Anonymous Walk Embeddings, each network flow graph is encoded into a latent representation that encapsulates both the structure of the graph and the features of its nodes, while minimizing information loss. This results in semantically-rich and robust representation vectors which can be manipulated by machine learning algorithms to perform downstream network-related tasks. To evaluate our network flow embedding models, we use probing flows captured with two /20 network telescopes and labeled using reports originating from different sources. The experimental results show that the proposed network flow embedding method allows for reliable darknet probing activity classification. Furthermore, a comparison between our self-supervised approach and a fully-supervised graph convolutional network shows that, in situations with limited labeled data, the downstream classification model that uses the derived latent representations as inputs outperforms the fully-supervised graph convolutional network. There are many applications of this research work in cybersecurity, such as network flow clustering, attack detection and prediction, malware detection, vulnerability exploit analysis, and inference of attacker’s intentions.
- 29 March 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 90749-90765
Open AccessSelf-Supervised Latent Representations of Network Flows and Application to Darknet Traffic Classification
Characterizing network flows is essential for security operators to enhance their awareness about cyber-threats targeting their networks. The automation of network flow characterization with machine learning has received much attention in recent years. To this aim, raw network flows need to be transformed into structured and exploitable data. In this research work, we propose a method to encode raw network flows into robust latent representations exploitable in downstream tasks. First, raw network flows are transformed into graph-structured objects capturing their topological aspects (packet-wise transitional patterns) and features (used protocols, packets’ flags, etc.). Then, using self-supervised learning techniques like Graph Auto-Encoders and Anonymous Walk Embeddings, each network flow graph is encoded into a latent representation that encapsulates both the structure of the graph and the features of its nodes, while minimizing information loss. This results in semantically-rich and robust representation vectors which can be manipulated by machine learning algorithms to perform downstream network-related tasks. To evaluate our network flow embedding models, we use probing flows captured with two /20 network telescopes and labeled using reports originating from different sources. The experimental results show that the proposed network flow embedding method allows for reliable darknet probing activity classification. Furthermore, a comparison between our self-supervised approach and a fully-supervised graph convolutional network shows that, in situations with limited labeled data, the downstream classification model that uses the derived latent representations as inputs outperforms the fully-supervised graph convolutional network. There are many applications of this research work in cybersecurity, such as network flow clustering, attack detection and prediction, malware detection, vulnerability exploit analysis, and inference of attacker’s intentions.
- 15 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network and Service Management
- Vol. 19 (4) , 5170-5182
Monitoring Network Telescopes and Inferring Anomalous Traffic Through the Prediction of Probing Rates
Network reconnaissance is the first step preceding a cyber-attack. Hence, monitoring the probing activities is imperative to help security practitioners enhancing their awareness about Internets large-scale events or peculiar events targeting their network. In this paper, we present a framework for an improved and efficient monitoring of the probing activities targeting network telescopes. Particularly, we model the probing rates which are a good indicator for measuring the cyber-security risk targeting network services. The approach consists of first inferring groups of network ports sharing similar probing characteristics through a new affinity metric capturing both temporal and semantic similarities between ports. Then, sequences of probing rates targeting similar ports are used as inputs to stacked Long Short-Term Memory (LSTM) neural networks to predict probing rates 1 hour and 1 day in advance. Finally, we describe two monitoring indicators that use the prediction models to infer anomalous probing traffic and to raise early threat warnings. We show that LSTM networks can accurately predict probing rates, outperforming the nonstationary autoregressive model, and we demonstrate that the monitoring indicators are efficient in assessing the cyber-security risk related to vulnerability disclosure.
- 15 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network and Service Management
- Vol. 19 (4) , 5170-5182
Monitoring Network Telescopes and Inferring Anomalous Traffic Through the Prediction of Probing Rates
Network reconnaissance is the first step preceding a cyber-attack. Hence, monitoring the probing activities is imperative to help security practitioners enhancing their awareness about Internets large-scale events or peculiar events targeting their network. In this paper, we present a framework for an improved and efficient monitoring of the probing activities targeting network telescopes. Particularly, we model the probing rates which are a good indicator for measuring the cyber-security risk targeting network services. The approach consists of first inferring groups of network ports sharing similar probing characteristics through a new affinity metric capturing both temporal and semantic similarities between ports. Then, sequences of probing rates targeting similar ports are used as inputs to stacked Long Short-Term Memory (LSTM) neural networks to predict probing rates 1 hour and 1 day in advance. Finally, we describe two monitoring indicators that use the prediction models to infer anomalous probing traffic and to raise early threat warnings. We show that LSTM networks can accurately predict probing rates, outperforming the nonstationary autoregressive model, and we demonstrate that the monitoring indicators are efficient in assessing the cyber-security risk related to vulnerability disclosure.
- Open Access
- 19 November 2021
- Published by MDPI AG in International Journal of Environmental Research and Public Health
- Vol. 18 (22) , 12172
Real-Time Infoveillance of Moroccan Social Media Users’ Sentiments towards the COVID-19 Pandemic and Its Management
The impact of COVID-19 on socio-economic fronts, public health related aspects and human interactions is undeniable. Amidst the social distancing protocols and the stay-at-home regulations imposed in several countries, citizens took to social media to cope with the emotional turmoil of the pandemic and respond to government issued regulations. In order to uncover the collective emotional response of Moroccan citizens to this pandemic and its effects, we use topic modeling to identify the most dominant COVID-19 related topics of interest amongst Moroccan social media users and sentiment/emotion analysis to gain insights into their reactions to various impactful events. The collected data consists of COVID-19 related comments posted on Twitter, Facebook and Youtube and on the websites of two popular online news outlets in Morocco (Hespress and Hibapress) throughout the year 2020. The comments are expressed in Moroccan Dialect (MD) or Modern Standard Arabic (MSA). To perform topic modeling and sentiment classification, we built a first Universal Language Model for the Moroccan Dialect (MD-ULM) using available corpora, which we have fine-tuned using our COVID-19 dataset. We show that our method significantly outperforms classical machine learning classification methods in Topic Modeling, Emotion Recognition and Polar Sentiment Analysis. To provide real-time infoveillance of these sentiments, we developed an online platform to automate the execution of the different processes, and in particular regular data collection. This platform is meant to be a decision-making assistance tool for COVID-19 mitigation and management in Morocco.
- 19 November 2021
- Published by MDPI AG in International Journal of Environmental Research and Public Health
- Vol. 18 (22) , 12172
Open AccessReal-Time Infoveillance of Moroccan Social Media Users’ Sentiments towards the COVID-19 Pandemic and Its Management
The impact of COVID-19 on socio-economic fronts, public health related aspects and human interactions is undeniable. Amidst the social distancing protocols and the stay-at-home regulations imposed in several countries, citizens took to social media to cope with the emotional turmoil of the pandemic and respond to government issued regulations. In order to uncover the collective emotional response of Moroccan citizens to this pandemic and its effects, we use topic modeling to identify the most dominant COVID-19 related topics of interest amongst Moroccan social media users and sentiment/emotion analysis to gain insights into their reactions to various impactful events. The collected data consists of COVID-19 related comments posted on Twitter, Facebook and Youtube and on the websites of two popular online news outlets in Morocco (Hespress and Hibapress) throughout the year 2020. The comments are expressed in Moroccan Dialect (MD) or Modern Standard Arabic (MSA). To perform topic modeling and sentiment classification, we built a first Universal Language Model for the Moroccan Dialect (MD-ULM) using available corpora, which we have fine-tuned using our COVID-19 dataset. We show that our method significantly outperforms classical machine learning classification methods in Topic Modeling, Emotion Recognition and Polar Sentiment Analysis. To provide real-time infoveillance of these sentiments, we developed an online platform to automate the execution of the different processes, and in particular regular data collection. This platform is meant to be a decision-making assistance tool for COVID-19 mitigation and management in Morocco.
- Open Access
- 27 June 2020
- Published by Springer Nature
- p. 45-54
ThreatPredict: From Global Social and Technical Big Data to Cyber Threat Forecast
- 27 June 2020
- Published by Springer Nature
- p. 45-54
Open AccessThreatPredict: From Global Social and Technical Big Data to Cyber Threat Forecast
- Open Access
Deep Learning for Inferring the Surface Solar Irradiance from Sky Imagery
We present a novel approach to perform ground-based estimation and prediction of the surface solar irradiance with the view to predicting photovoltaic energy production. We propose the use of mini-batch k-means clustering to extract features, referred to as per cluster number of pixels (PCNP), from sky images taken by a low-cost fish eye camera. These features are first used to classify the sky as clear or cloudy using a single hidden layer neural network; the classification accuracy achieves 99.7%. If the sky is classified as cloudy, we propose to use a deep neural network having as input features the PCNP to predict intra-hour variability of the solar irradiance. Toward this objective, in this paper, we focus on estimating the deep neural network model relating the PCNP features and the solar irradiance, which is an important step before performing the prediction task. The proposed deep learning-based estimation approach is shown to have an accuracy of 95%.
Open AccessDeep Learning for Inferring the Surface Solar Irradiance from Sky Imagery
We present a novel approach to perform ground-based estimation and prediction of the surface solar irradiance with the view to predicting photovoltaic energy production. We propose the use of mini-batch k-means clustering to extract features, referred to as per cluster number of pixels (PCNP), from sky images taken by a low-cost fish eye camera. These features are first used to classify the sky as clear or cloudy using a single hidden layer neural network; the classification accuracy achieves 99.7%. If the sky is classified as cloudy, we propose to use a deep neural network having as input features the PCNP to predict intra-hour variability of the solar irradiance. Toward this objective, in this paper, we focus on estimating the deep neural network model relating the PCNP features and the solar irradiance, which is an important step before performing the prediction task. The proposed deep learning-based estimation approach is shown to have an accuracy of 95%.