IFIP Open Access Digital Library
Not a member yet
22614 research outputs found
Sort by
Digital Detectives: A Serious Point-and-Click Game for Digital Forensics
No full textPart 3: Digital Forensics and InvestigationInternational audienceAs cyberattacks increase yearly, the ongoing scarcity of cybersecurity professionals and lack of knowledge about digital forensics (DF) are substantial challenges for organizations. One potential solution to mitigate the severe implications of cyberattacks is to increase DF. This work introduces a serious game called Digital Detectives, designed to educate students about DF activities and raise awareness for non-specialist employees. Players take on the role of a DF investigator tasked with probing a cyberattack on a fictional company. Players acquire practical knowledge, awareness, and skills through gamified evidence collection, attacker identification, tool usage, and data recovery. The game features AI-generated graphics and enables players to explore various places to gather evidence. Collecting items and engaging with mini-games, such as analyzing network activity, immerses learners in an interactive setting. The serious game’s objective is to promote an understanding of the DF profession, thereby aiding in future investigations and cultivating awareness of behaviors that could contribute to security breaches. We evaluated the serious game with 36 students and found a significant increase in knowledge about DF and cybersecurity awareness. Simply put, Digital Detectives is a valuable tool to educate students in forensics and raise awareness for employees
Dependency-Type Weighted Graph Convolutional Network on End-to-End Aspect-Based Sentiment Analysis
No full textPart 1: Pattern RecognitionInternational audiencePrevious studies consider little on using dependency-type messages in the E2E-ABSA task. Studies using dependency-type messages just contact the dependency-type message and word embedding vectors, which may not fully fuse the context feature and information from the dependency type. This paper proposes a new model called Dependency-Type Weighted Graph Convolution Network (DTW-GCN) to compose dependency-type messages and word embedding. We use a type-weighted matrix to combine the dependency-type message, and DTW-GCN could fuse the dependency-type message and word embedding vectors. Experiments conducted on three benchmark datasets verify the effectiveness of our model
A Concept-Based Local Interpretable Model-Agnostic Explanation Approach for Deep Neural Networks in Image Classification
No full textPart 2: Image UnderstandingInternational audienceA well-recognized and widely-used explainable artificial intelligence (XAI) method is Local Interpretable Model-agnostic Explanations (LIME), which offers instance-level interpretation by generating new data around the instance and training a locally interpretable linear model. However, when using LIME to explain the image classification model, it generates interpretations at the level of super-pixel representation. This does not assure comprehensibility to humans due to the lack of semantic information in super-pixels. To enhance the intelligibility of LIME, we propose an advanced version of LIME, termed Concept-based Local Interpretable Model-agnostic Explanations (ConceptLIME). In ConceptLIME, the explanations are formulated in terms of human-understandable concepts as opposed to the semantically deficient super-pixels, thereby augmenting the comprehensibility of the original LIME method. Comparative experiments have been conducted between ConceptLIME and LIME to validate the effectiveness of ConceptLIME. The experimental results indicate that ConceptLIME outperforms LIME regarding predictive performance on both the perturbation dataset and the explained instances. Moreover, the fidelity of the explanations generated by ConceptLIME surpasses that produced by LIME. The interpretations provided by ConceptLIME are more intelligible and intuitive than LIME’s explanations. Consequently, our proposed ConceptLIME exhibits superior properties, including predictive performance, fidelity, and comprehensibility, when compared with LIME
Three-Dimensional Bin Packing Problems with the Operating Time of a Robot Manipulator
No full textPart 1: Smart and Sustainable Supply Chain Management in the Society 5.0 EraInternational audienceMany optimization algorithms for solving three-dimensional bin-packing problems have been addressed. The problem is to minimize the cost of the bins in which packages are placed with the aim of reducing transportation costs and improving productivity. From the viewpoint of actual optimization of the entire factory, it is necessary to consider the motion planning of a robot manipulator to handle the items in the three-dimensional bin packing problem. This study considers a three-dimensional robotic bin packing problem, which minimizes the weighted sum of bin cost and the robot’s operating time. A genetic algorithm is developed using Sequence-Triple representation, a layout representation of a three-dimensional arrangement. For robot motion planning, the Rapidly Exploring Random Tree Star is used for trajectory generation to obtain near-optimal solutions. Computational experiments show the superiority of the proposed optimization method by comparing it with a conventional sequential optimization method
Integrating Ontology with Cobot Execution for Human-Robot Collaborative Assembly Using Heterogenous Cobots
No full textPart 2: Human-centred Manufacturing and Logistics Systems Design and Management for the Operator 5.0International audienceThe manufacturing industry has heavily relied on robotic automation since the third industrial revolution. However, traditional robots struggle to adapt to the variability of modern production demands. Collaborative robots (cobots) offer a solution by providing flexible automation tailored to personalized manufacturing requirements. Human-robot collaboration (HRC) systems, particularly beneficial for small-scale manufacturing, integrate human capabilities with technological advancements. Cobots enhance productivity, offer ergonomic benefits, and facilitate automation transformation through connectivity and data analytics. Effective communication between resources is crucial for enabling the execution of shared subtasks and ensuring the safety and efficiency of collaborative assembly. This study proposes integrating ontology with cobot execution programs to enhance collaborative assembly operations. The ontology encompasses knowledge related to the products to be assembled, the HRC environment (including resources, tools, and regions), and the relations between these entities and monitoring data. Additionally, establishing a connection between cobot controllers enables the seamless exchange of commands for coordinated execution. By enabling communication between resources, collaborative assembly tasks can be executed either sequentially or simultaneously, monitored for progress, and adjusted as necessary without manual intervention. The proposed approach is applied in the human-robot collaborative mold assembly for cobot execution. The integration of ontology simplifies cobot programming, modifications according to order changes, and facilitates cobot execution control based on monitoring results. This study contributes to streamline cobot execution and decision-making processes in manufacturing environments by incorporating ontology-based knowledge into cobot execution programs
Renal Irregularities Detection Using Convolutional Neural Network
No full textPart 1: SDG 3 Good Health and Well-BeingInternational audienceRenal irregularities are serious medicinal condition that is becoming more common and killing more people each year. In its early stages, renal irregularities are curable, but it can progress irreversibly and result in renal failure. Cyst development, kidney tumours, and stone are the three predominant kidney irregularities that impair renal function among numerous diseases. The prompt detection and treatment of renal disease is a major challenge to the medical community. If renal problems like stones, cysts and tumours are not detected at an early stage, renal failure may ensue. Computer-assisted diagnostics is necessary to complement medical assessments made by clinicians and specialists as renal disease is more spread, there are less clinicians accessible, assessment and monitoring rates are rising, especially in developing nations. Though they still don’t perform well, artificial intelligence methods like machine and deep learning have been utilized in literature to identify illness. This study uses CNN model for renal disease categorization and prognosis that is based on deep learning. For exploration, we use a benchmark CT kidney dataset from Computed Tomography. CNN first preprocesses the data before extracting the features from the pictures. The suggested method successfully classifies renal illness, with a notable accuracy of 99.3%, 99.5% precision, 95.3% recall, and 9.88% F1-score
A Framework for Matching Distinct Personality Types with Information Security Awareness Methods
No full textPart 1: Awareness and EducationInternational audienceThe objective of this study is to develop a framework to associate learning styles and social influencing vulnerabilities with different personality types in the context of tailoring Information Security Awareness (ISA) methods for people with different personality types. Directed content analysis is carried out to develop the framework. The analysis is conducted in the following two parts: a). Describe and identify keywords for the DISC (Dominance (D), Inducement (I), Submission (S) and Compliance (C)) personality types, Kolb’s learning styles and Cialdini’s social influencing principles; b). Identify the relationships between Personality types, Learning styles, and Social influencing vulnerabilities and create the PLS (i.e., Personality types, Learning styles, and Social influencing vulnerabilities) framework. As a result, four relationships are identified for each distinct personality type in the PLS framework. This study contributes to building a sound theoretical ground for tailoring ISA methods for people with different personality types. In addition, the derived keywords are helpful to capture a good understanding of the different dimensions of the selected theories. Furthermore, the developed PLS framework can be used as a base for managers to employ ISA methods for people with different personality types in organizations
Towards an Active Learning Approach for the Design of a Secure Programming Course Using Constructive Alignment
No full textPart 1: Awareness and EducationInternational audienceEven though economic development encourages innovation across the world, the education industry has remained relatively consistent in following traditional modes of teaching. Traditional modes of teaching include the instructor transferring content to students through presenting the content either in a classroom setting, or through online means, often with little attention given to active learning and engagement of students. This form of education may not be effective particularly in Science, Technology, Engineering, and Mathematics (STEM) subjects, such as programming, which is popular in STEM related qualifications. The traditional teaching mode often negatively affects students’ success rates resulting in unemployment. Due to the passiveness of the traditional teaching modes, students often lose concentration during lengthy lectures. Therefore, a more active learning approach is recommended due to its engaging and effective mode of teaching. This paper investigates the relevant active learning elements which could assist in effective secure programming education. The identified elements are used to design a secure programming course within a constructive alignment approach thereby ensuring that intended learning outcomes are achieved
Scoping Review: The Landscape of Digital Risks and Cybersecurity Solutions for Journalists
No full textPart 1: Management and RiskInternational audienceThe ever-evolving digital world impacts those far beyond the Information Systems (IS) field. Due to the nature of current day journalism practices, journalists are required to use digital resources to be successful in their line of work. Meanwhile, they are battling new and existing digital risks that target their work and personal lives and can lead to psychological and physical harm, reputational damage, and diminished freedom of the press. This article identifies the current landscape of digital risks that individual journalists face in their line of work to address areas in which cybersecurity and journalism can intersect. Based on our findings, online harassment and surveillance are the most identified and researched digital risks to journalists. Therefore, future IS security research could help further identify risks outside these two domains. A comparison is then presented between journalists’ digital risk management recommendations and prominent cybersecurity risk management guidelines, including ISO 31000, NIST frameworks, and a Privacy Impact Assessment (PIA). The most prominent finding from the comparison is that cybersecurity risk management guidelines are process based while journalists risk management recommendations focus on specific action items. Lastly, the paper addresses whether cybersecurity and journalists risk management practices provide individual journalists with a comprehensive process to address their digital risks
A Systematic Review of Various Deep Learning Techniques for Network Intrusion Detection System
No full textPart 1: Applications of AI/ML in KDM, Cloud Computing & SecurityInternational audienceThe collective frequency and intricacy of attacks on computer networks remain a threat to information security within computer schemes. In order to solve this, researchers exploited network intrusion detection systems (NIDSs) to safe guard networks and information. Because of its lively nature of malware and continuous change in attacks, these systems normally recognize intrusions by the examination of network traffic. This survey demonstrates several Deep Learning (DL) approaches for the identification and categorization of unexpected attacks. The constant variations in the activities of the network make it essential to examine several datasets by dynamic and static techniques. Here 25 research papers are analyzed and surveyed. A complete assessment of many DL classifiers was exposed on various benchmark malware datasets. First, it offers a classification of computer intrusions with an explanation of categorized methods. Next, a general outline of the NIDS is described with its basic features. Then, the classification of NIDS based on their categorized methods is explained. After that, the challenges faced by the categorized methods are deliberated in the section of research gaps and issues. Lastly, the analysis in the survey is performed in terms of publication year, category analysis, tools used and performance metrics