IFIP Open Access Digital Library
Not a member yet
22614 research outputs found
Sort by
Unlocking User Privacy: A Systematic Survey of Factors and Methods in Predicting App Permission Decisions
No full textPart 2: Machine LearningInternational audienceMobile devices have become an indispensable part of everyday life for most people and have improved our lives in many ways, offering a multitude of possibilities through the various available applications. On one hand, the risks around information privacy on mobile devices become more and more challenging. On the other hand, users are called to analyse and take plenty of complex privacy decisions. There have been many attempts to overcome this burden from users by automating the process of granting permissions to ap-plications. This study provides a literature review on works for predicting users’ privacy decisions in application requests. Our research aims to shed light on the different factors that have been identified in the literature as important predictors of users’ decisions on app requests, as well as the methods that have been used, focusing on machine learning approaches and building user privacy profiles
Malicious Insider Threat Detection Using Sentiment Analysis of Social Media Topics
No full textInternational audienceMalicious insiders often pose a danger to information security systems, which can be a crucial challenge to tackle. Existing technological solutions attempt to identify potential threats via their anomalous system interactions, however, fully fail to suppress the rise in costly data breaches, initiated by trusted users who exploit their authorised access for unauthorised means. Although alternative proposals incorporate a psychosocial angle by utilising correlations between real-world insider cases and their emotional state, personality type or predispositions, they also pose several limitations. In order to mitigate the challenges, this work builds on such profiling methodologies but directly harnesses language as a behavioural indicator, by applying the Natural Language Processing technique of sentiment analysis. It offers a novel approach to lowering the risk of potential insiders and thus taking advantage of the wealth of discourse made public by social media sites to focus on one trait of the narcissist, lack of empathy, and another with a negative correlation with narcissism and compassion. It demonstrates how the careful choice of social media topics can act as a catalyst for language indicating low levels of empathy and compassion, and facilitating the detection of malicious insiders, via their proven tendency towards narcissism
Neurosymbolic Learning in the XAI Framework for Enhanced Cyberattack Detection with Expert Knowledge Integration
No full textInternational audienceThe perpetual evolution of cyberattacks, especially in the realm of Internet of Things (IoT) networks, necessitates advanced, adaptive, and intelligent defence mechanisms. The integration of expert knowledge can drastically enhance the efficacy of IoT network attack detection systems by enabling them to leverage domain-specific insights. This paper introduces a novel approach by applying Neurosymbolic Learning within the Explainable Artificial Intelligence (XAI) framework to enhance the detection of IoT network attacks while ensuring interpretability and transparency in decision-making. Neurosymbolic Learning synergizes symbolic AI, which excels in handling structured knowledge and providing explainability, with neural networks, known for their prowess in learning from data. Our proposed model utilizes expert knowledge in the form of rules and heuristics, integrating them into a learning mechanism to enhance its predictive capabilities and facilitate the incorporation of domain-specific insights into the learning process. The XAI framework is deployed to ensure that the predictive model is not a “black box”, providing clear, understandable explanations for its predictions, thereby augmenting trust and facilitating further enhancement by domain experts. Through rigorous evaluation against benchmark IoT network attack datasets, our model demonstrates superior detection performance compared to prevailing models, along with enhanced explainability and the successful incorporation of expert knowledge into the adaptive learning process. The proposed approach not only fortifies the security mechanisms against network attacks in IoT environments but also ensures that the knowledge discovery and decision-making processes are transparent, interpretable, and verifiable by human experts
Identification of Cyber Threats and Vulnerabilities in Norwegian Distribution Networks
No full textInternational audienceThis paper presents cyber threats and vulnerabilities in Norwegian power distribution networks identified from historical incidents and practical experiences over the last decade
Image-Based Human Action Recognition with Transfer Learning Using Grad-CAM for Visualization
No full textPart 1: Biomedical/ClassificationInternational audienceHuman action recognition (HAR) in still images is a critical task for various applications ranging from surveillance to human-computer interaction. This research introduces an innovative approach to HAR using transfer learning. By utilizing the InceptionResNetV2 architecture, pre-trained on ImageNet, we fine-tuned a model on the Data Sprint 76 dataset from AIPlanet. This dataset features a constrained set of still images depicting 15 human actions. Our objective was to leverage the pre-trained network’s rich feature extraction capabilities and adapt them to the domain-specific task of HAR. We implemented a two-phase training process, initially training the custom model with frozen base layers to learn the new classes, followed by selectively unfreezing one-fourth and fine-tuning the base model to improve the feature adaptation to the HAR task. Data augmentation techniques were crucial in simulating a more extensive dataset, mitigating the overfitting risk, and enhancing the model’s generalization abilities. The effectiveness of our model was quantified by a training accuracy of 88.43% and a validation accuracy of 77.30%. To interpret the model’s decision-making process, we integrated Gradient-weighted Class Activation Mapping (Grad-CAM), which provided visual explanations for the model’s predictions. This insight was critical in understanding the model’s focus areas within the images and aided in identifying the cause of misclassifications, as observed in the confusion matrix. This study demonstrates that transfer learning, coupled with advanced visualization techniques like Grad-CAM, can effectively mitigate the challenges posed by limited data in HAR
Fake News in Developing Countries: Drivers, Mechanisms and Consequences
No full textPart 2: Information and Computer SecurityInternational audienceFake news has become a global problem. Drivers in the spread of fake news include the growth of the world wide web and the growth in the use of smartphones and social media. In the field of ICT4D, access to mobile phones and the internet has strongly been promoted in developing countries, for the sake of socio-economic development. However along with the availability and use of mobile technology, the Internet and social media in developing countries, the prevalence and risk of fake news have also increased. While many studies have been performed on fake news, most of these have been done in the Global North. This study attempts to address this gap by investigating the state of knowledge of fake news in the Global South. A systematic literature review is performed that focuses on the spreading and consequences of fake news in developing countries. A thematic analysis found that, like in the Global North, fake news in the Global South is predominantly spread by means of social media. The consequences of the spreading of fake news in developing countries include an increased mistrust in mainstream media, as well as a mistrust in various vaccines which leads to a health risk. Further, fake news has been shown to contribute to violent unrest and to worsen religious crises. Fake news poses the risk of undermining the benefits associated with increased connectivity in developing countries. This is a call for the ICT4D community to consider ways to address the risk of fake news
An Indicator Based Evolutionary Algorithm for Multiparty Multiobjective Knapsack Problems
No full textPart 3: Neural and Evolutionary ComputingInternational audienceAs a special case of the multiobjective optimization problem, the multiobjective knapsack problem (MOKP) widely exists in real-world applications. Currently, most algorithms used to solve MOKPs assume that these problems involve only one decision maker (DM). However, some complex MOKPs often involve more than one decision makers and we call such problems multiparty multiobjective knapsack problems (MPMOKPs). Existing algorithms cannot solve MPMOKPs effectively. To the best of our knowledge, there is only a little attention paid to MPMOKPs. In this paper, inspired by existing SMS-EMOA, we propose a novel indicator-based algorithm called SMS-MPEMOA to solve MPMOKPs, which aims to search solutions to satisfy all decision makers as much as possible. SMS-MPEMOA is compared with several state-of-the-art multiparty multiobjective optimization algorithms (MPMOEAs) on the benchmarks and the experimental results demonstrate that SMS-MPEMOA is very competitive
Hybrid Integrated Dimensionality Reduction Method Based on Conformal Homeomorphism Mapping
No full textPart 1: Machine LearningInternational audienceBased on the theories of Riemannian surface, Topology and Analytic function, a novel method for dimensionality reduction is proposed in this paper. This approach utilizes FCA to merge highly correlated features to obtain approximate independent new features in the locally, and establishes a conformal homomorphic function to realize global dimensionality reduction for text data with the manifold embed in the Hausdorff space. During the process of dimensionality reduction, the geometric topological structure information of the original data is preserved through conformal homomorphism function. This method is characterized by its simplicity, effectiveness, low complexity, and it avoids the neighbor problem in nonlinear dimensionality reduction and it is conducive to the outlier data. Moreover, it has extensible for new text vectors and new feature from sub-vectors of new text vectors, and incremental operation without involving existing documents. The mapping function exhibits desirable properties resulting in stable, reliable, and interpretable dimensionality reduction outcomes. Experimental results on both construction laws and regulations dataset and toutiao text dataset demonstrate that this dimensionality reduction technique is effective when combined with the typical classification method of Random Forest, Support Vector Machine, and Feedforward Neural Network
Entropy-Based Logic Explanations of Differentiable Decision Tree
No full textPart 1: Machine LearningInternational audienceExplainable reinforcement learning has evolved rapidly over the years because transparency of the model’s decision-making process is crucial in some important domains. Differentiable decision trees have been applied to this field due to their performance and interpretability. However, the number of parameters per branch node of a differentiable decision tree is related to the state dimension. When the feature dimension of states increases, the number of states considered by the model in each branch node decision also increases linearly, which increases the difficulty of human understanding. This paper proposes a entroy-based differentiable decision tree, which can restrict each branch node to use as few features as possible to predict during the training process. After the training is completed, the parameters that have little impact on the output of the branch node will be blocked, thus significantly reducing the decision complexity of each branch node. Experiments in multiple environments demonstrate the significant interpretability advantage of our proposed approach
Efficient and Secure Authentication Scheme for Internet of Vehicles
No full textPart 5: Business Intelligence and Risk ControlInternational audienceThe Internet of Vehicles (IoV) improves efficiency of transportation systems while enhancing the passenger travel experience. However, due to the open wireless communication environment, the IoV requires a reliable and secure authentication and key agreement scheme to ensure that the exchanged data in public channel cannot be forged or modified by the adversary. In most existing authentication schemes, the vehicle usually authenticates with each other by an online Trusted Authority (TA), which results in the authentication efficiency of these centralized authentication schemes are easily affected by TA’s computational and communication bottlenecks as the increase of traffic density. Therefore, this paper proposes a secure and efficient authentication and key agreement scheme for IoV, where the vehicles can authenticate with each other and build a session key through a pre-shared key. Besides, a group key is also proposed to broadcast basic safety messages in the same RSU group securely. The group key can be updated when the vehicle joins and leaves, so that a leaving group member cannot access the current communication process. By the Heuristic and BAN logic analysis, the proposed scheme is proved to be secure. Compared with existing schemes, the proposed scheme meets the security requirements and has significant advantages in terms of computation and communication overhead