Publications tagged with ITS applications
Published:
Publications tagged with "ITS applications"
- DetailsCampanile, L., De Fazio, R., Di Giovanni, M., & Marulli, F. (2024). Beyond the Hype: Toward a Concrete Adoption of the Fair and Responsible Use of AI [Conference paper]. CEUR Workshop Proceedings, 3762, 60–65. https://www.scopus.com/inward/record.uri?eid=2-s2.0-85205601768&partnerID=40&md5=99140624de79e37b370ed4cf816c24e7
Abstract
Artificial Intelligence (AI) is a fast-changing technology that is having a profound impact on our society, from education to industry. Its applications cover a wide range of areas, such as medicine, military, engineering and research. The emergence of AI and Generative AI have significant potential to transform society, but they also raise concerns about transparency, privacy, ownership, fair use, reliability, and ethical considerations. The Generative AI adds complexity to the existing problems of AI due to its ability to create machine-generated data that is barely distinguishable from human-generated data. Bringing to the forefront the issue of responsible and fair use of AI. The security, safety and privacy implications are enormous, and the risks associated with inappropriate use of these technologies are real. Although some governments, such as the European Union and the United States, have begun to address the problem with recommendations and proposed regulations, it is probably not enough. Regulatory compliance should be seen as a starting point in a continuous process of improving the ethical procedures and privacy risk assessment of AI systems. The need to have a baseline to manage the process of creating an AI system even from an ethics and privacy perspective becomes progressively more important In this study, we discuss the ethical implications of these advances and propose a conceptual framework for the responsible, fair, and safe use of AI. © 2024 Copyright for this paper by its authors. - DetailsMarulli, F., Verde, L., & Campanile, L. (2021). Exploring data and model poisoning attacks to deep learning-based NLP systems [Conference paper]. Procedia Computer Science, 192, 3570–3579. https://doi.org/10.1016/j.procs.2021.09.130
Abstract
Natural Language Processing (NLP) is being recently explored also to its application in supporting malicious activities and objects detection. Furthermore, NLP and Deep Learning have become targets of malicious attacks too. Very recent researches evidenced that adversarial attacks are able to affect also NLP tasks, in addition to the more popular adversarial attacks on deep learning systems for image processing tasks. More precisely, while small perturbations applied to the data set adopted for training typical NLP tasks (e.g., Part-of-Speech Tagging, Named Entity Recognition, etc..) could be easily recognized, models poisoning, performed by the means of altered data models, typically provided in the transfer learning phase to a deep neural networks (e.g., poisoning attacks by word embeddings), are harder to be detected. In this work, we preliminary explore the effectiveness of a poisoned word embeddings attack aimed at a deep neural network trained to accomplish a Named Entity Recognition (NER) task. By adopting the NER case study, we aimed to analyze the severity of such a kind of attack to accuracy in recognizing the right classes for the given entities. Finally, this study represents a preliminary step to assess the impact and the vulnerabilities of some NLP systems we adopt in our research activities, and further investigating some potential mitigation strategies, in order to make these systems more resilient to data and models poisoning attacks. © 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of KES International.