IFIP Open Access Digital Library
Not a member yet
22614 research outputs found
Sort by
User Perceptions of CAPTCHAs: University vs. Internet Users
No full textPart 6: Security User StudiesInternational audienceWe surveyed over 250 participants from a university and Amazon Mechanical Turk to understand user perceptions of CAPTCHA security and usability. Users reported struggling with CAPTCHA challenges due to increasing difficulty, leading to frustration and a negative experience. Concerns about reliability and security were also noted. Our findings offer valuable insights for developing more secure and user-friendly CAPTCHA technologies
Public Management Competencies in a Digital World: Lessons from a Global Frontrunner
No full textInternational audienceIn contemporary public administration, mid-level managers play a key role in shaping if and how digital technology is used on every-day basis. But knowledge of the competencies of mid-level managers who increasingly need to manage use of digital technology is rather limited. Seeking to rectify the lack of knowledge, we distil relevant mid-level managerial competencies within thinking on bureaucracies, digital transformation and “digital skills”, and combine this with a multiple case-study of Danish administrative bodies’ use of automated decision-making. Drawing on an understanding of competences as a combination of work-related knowledge, skills and aptitudes held by individuals, we tentatively identify five key competencies for mid-level managers in relation to public administrative bodies’ everyday use of digital technology. The identified competencies can help individual managers and public administrative bodies being more targeted in their efforts to manage increased technology use on an every-day basis. Noteworthy, none of the identified competencies imply managers should become technology specialists today or anytime in the future
Building a Responsible AI Ecosystem: Thailand’s Journey Towards Ethical AI
No full textInternational audienceAI Thailand is a national program aiming to prepare essential infrastructure for artificial intelligence (AI) development in Thailand to promote economic growth and increase the country’s competitiveness. Multiple facets of infrastructure, human capacity, and a practical ecosystem need to be established. As a result, Thailand’s national AI strategy and action plan (2022–2027), known as NAIS, was approved by the Prime Minister’s Cabinet Office on July 26, 2022. The strategy aims to provide Thailand with an effective ecosystem for AI development. Accordingly, by implementing its strategy, NAIS will concentrate on enhancing the economy and improving Thai people’s quality of life by 2027. There are five strategies and fifteen work plans, and this work pays attention to the first strategy that prepares Thailand’s readiness in social, ethical, law, and regulation for AI applications
The History of AI in Thailand: Thickening Our Vision of AI by Caring for a Marginalised Actor
No full textInternational audienceThe first steps of artificial intelligence (AI) in Thailand can be traced back to Thai university course material from 1975. Around the mid-1980s, the Thai government started to acknowledge the potential of AI technology to become a driver for economic development. Yet, despite the benefits of monitoring AI’s domestic developments, such research remains scarce for Thailand. This paper seeks to fill this current literature gap through offering a centralised compilation of information about the unfolding of AI technology in Thailand from its beginnings to this day. Given the scarcity in written forms of publicly available knowledge, we enrich the available literature with three qualitative expert interviews, and we contextualise Thailand’s case within the broader Southeast Asian and Asian context. Thailand is regularly overlooked or marginalised when we ‘think AI’. However, we will understand that Thailand’s role in AI’s history is critical if willing to ‘thicken’ existing descriptions of AI. Reductivist understandings of AI might speak of a technology transfer of the 1970s from the Occident to Thailand and stop there. However, firstly, this obscures China’s growing influence over Thailand’s AI development since 2015. Secondly, this obscures Thailand’s necessarily participative role in AI’s development—both within Thailand and abroad—which thus contrasts with the understanding where Thailand would purely be subjected to ‘foreign’ influence in this technology transfer. Reviewing the history of AI in Thailand allows to add critically missing layers to our currently too partial understanding of the history of AI globally. We also hope to insufflate new air into this musty research avenue to further improve our understanding of Thailand’s contemporary sociotechnical unfoldment
Navigating Trust: A Morphological Analysisnof Blockchain-as-a-Service Providers
No full textPart 9: Trust and Trustworthy Technologies in Collaborative NetworksInternational audienceIn various use cases, the blockchain manages to increase trust among participants in the supply chain. Whether it is the origin of critical raw materials or compliance with industry-speciï¬c requirements. The majority of companies utilize blockchain as a service providers for this purpose to beneï¬t from the advantages of this technology. Existing classiï¬cations in the literature often do not consider the decision-makers’ perspective, which can lead to an incomplete evaluation of the available options. We offer a morphological approach that identiï¬es relevant decision criteria for trust and security. This enables organizations to make informed decisions and increase conï¬dence in their business processes without having to invest their own resources in development and operations
Human-Centered Solutions Based on Automated Visual Inspection System
No full textPart 5: Zero Defects and Zero Waste Strategies in Industrial Collaborative NetworksInternational audienceConventional painting inspections in the wind industry rely on human labour to visually inspect the parts, and the mental and physical fatigue of the operator influences the accuracy. The advances introduced on camera sensors, robots, and artificial intelligence algorithms can create a collaborative working environment between operators and automated visual inspection system.One of the main challenges for the automatic painting inspection system is the integration with autonomous guided vehicle to perform the inspection routes, and the communication of the data generated during the inspection (defect detection, location, etc.) to the augmented reality platform to send the specific information to the operator to perform the painting repairing operation.Current research aims to define a framework where integrating non-destructive inspection technologies, autonomous guided vehicles, and augmented reality will allow a symbiotic collaboration between humans and robots. The interoperability between these three solutions will increase the Human-Robot Collaboration in a wind tower painting process, improving product quality assurance
Artificial Intelligence Applications and Innovations: 20th IFIP WG 12.5 International Conference, AIAI 2024, Corfu, Greece, June 27–30, 2024, Proceedings, Part III
No full textInternational audienceBook Front Matter of AICT 71
Test Case Features as Hyper-heuristics for Inductive Programming
No full textPart 3: Data Mining/ModelingInternational audienceInstruction subsets are heuristics that can reduce the size of the inductive programming search space by tens of orders of magnitude. Comprising many overlapping subsets of different sizes, they serve as predictions of the instructions required to code a solution for any problem. Currently, this approach employs a single, large family of subsets meaning that some problems can search thousands of subsets before a solution is found. In this paper we introduce the use of test case type signatures as hyper-heuristics to select one of many, smaller families of instruction subsets. The type signature for any set of test cases maps directly to a single family and smaller families mean that fewer subsets need to be considered for most problems. Having many families also permits subsets to be reordered to better reflect their relative occurrence in human code – again reducing the search space size for many problems. Overall the new approach can further reduce the size of the inductive programming search space by between 1 and 3 orders of magnitude, depending on the type signature. Larger and more consistent reductions are possible through the use of more sophisticated type systems. The potential use of additional test case features as hyper-heuristics and some other possible future work is also briefly discussed
Forecasting Longitudinal Acceleration in Urban Vehicles
No full textPart 1: Deep LearningInternational audienceIn the automotive industry, the accuracy and processing of sensor data are crucial for ensuring vehicle safety. Among various vehicle dynamics parameters, longitudinal acceleration stands out as a crucial parameter, as its accurate prediction is essential for safety mechanisms such as the anti-lock braking system (ABS) and the activation of airbags. This article presents a Long Short-Term Memory (LSTM) neural network model designed to estimate the future longitudinal acceleration of a vehicle, with a focus on minimizing the hardware resources needed for computational operations. The development of the model was done using a large quantity of experimental data. The model’s input consists of the past and current values of the speeds of the four wheels, lateral and longitudinal accelerations, the steering angle of the front axle, and the yaw rate. Our test results demonstrate that the developed model can forecast longitudinal acceleration 0.1 s ahead by utilizing the preceding 0.1 s of sensor data with high accuracy. Comparing our results with those published in the literature, we can conclude that the model we developed provides significantly more accurate predictions for longitudinal acceleration, achieved through the application of a less complex model
Explanations for Core Decomposition
No full textPart 2: Graphs/Neural Networks/Machine LearningInternational audienceThe problem of core decomposition in networks is fundamental in understanding their structural organization. Core decomposition methods have been applied to different problems and various scientific fields ranging from Social Network Analysis (community detection) to Epidemiology and Disease Spread (identification of core groups where transmission rates are high). This study delves into the causal explanations underlying core decomposition, with ultimate aim to unravel the intricate mechanisms driving the emergence and evolution of cores within networks.The causal explanations are based on the interventional notion of causality as introduced by Halpern and Pearl. Leveraging techniques developed previously by some authors, we pinpoint the causal relationships between network connectivity and core decomposition. To this end, at first we define what a causal explanation is, according to the core decomposition problem. Then, apart from the algorithmic contribution in identifying the causal explanations, we also provide different related experimental results that demonstrate its use