Papers by Sophia Karagiorgou
Zenodo (CERN European Organization for Nuclear Research), Jun 6, 2023
2 Can be left void 3 GFT QA Matteo Falsetta REVISION HISTORY Version Date Partner(s) Description ... more 2 Can be left void 3 GFT QA Matteo Falsetta REVISION HISTORY Version Date Partner(s) Description 0.10 2022-11-30 All Internal delivery to accomplish MS01 0.11 2023-03-06 LXS Updated ToC 0.12 2023-03-24 All First round of requirements 0.13 2023-03-30 All Second round of requirements 0.14 2023-04-07 All Third round of requirements 1.00 2023-04-10 LXS Submitted for internal review 1.10 2023-04-12 ULB Internal review 1.11 2023-04-18 UBI Internal review 2.00 2023-04-20 LXS Submitted for internal QA 2.1 2023-04-25 UPRC QA 3.00 2023-04-27 LXS Final submission Disclaimer This document contains information which is proprietary to the MobiSpaces Consortium. Neither this document nor the information contained herein shall be used, duplicated, or communicated by any means to a third party, in whole or parts, except with the prior consent of the MobiSpaces Consortium.
Privacy-preserving Data Federation for Trainable, Queryable and Actionable Data
arXiv (Cornell University), Jul 11, 2018
The tremendous growth of positioning technologies and GPS enabled devices has produced huge volum... more The tremendous growth of positioning technologies and GPS enabled devices has produced huge volumes of tracking data during the recent years. This source of information constitutes a rich input for data analytics processes, either offline (e.g. cluster analysis, hot motion discovery) or online (e.g. short-term forecasting of forthcoming positions). This paper focuses on predictive analytics for moving objects (could be pedestrians, cars, vessels, planes, animals, etc.) and surveys the state-of-the-art in the context of future location and trajectory prediction. We provide an extensive review of over 50 works, also proposing a novel taxonomy of predictive algorithms over moving objects. We also list the properties of several real datasets used in the past for validation purposes of those works and, motivated by this, we discuss challenges that arise in the transition from conventional to Big Data applications.
Sensors, Sep 13, 2022
This article is an open access article distributed under the terms and conditions of the Creative... more This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY
Sensors
Adversarial machine learning (AML) is a class of data manipulation techniques that cause alterati... more Adversarial machine learning (AML) is a class of data manipulation techniques that cause alterations in the behavior of artificial intelligence (AI) systems while going unnoticed by humans. These alterations can cause serious vulnerabilities to mission-critical AI-enabled applications. This work introduces an AI architecture augmented with adversarial examples and defense algorithms to safeguard, secure, and make more reliable AI systems. This can be conducted by robustifying deep neural network (DNN) classifiers and explicitly focusing on the specific case of convolutional neural networks (CNNs) used in non-trivial manufacturing environments prone to noise, vibrations, and errors when capturing and transferring data. The proposed architecture enables the imitation of the interplay between the attacker and a defender based on the deployment and cross-evaluation of adversarial and defense strategies. The AI architecture enables (i) the creation and usage of adversarial examples in th...
BigDataStack delivers a complete high-performant stack of technologies addressing the needs of da... more BigDataStack delivers a complete high-performant stack of technologies addressing the needs of data operations and applications. BigDataStack's holistic solution incorporates approaches for data-focused application analysis and dimensioning, and process modelling towards increased performance, agility and efficiency. A toolkit allowing the specification of analytics tasks in a declarative way, their integration in the data path, as well as an adaptive visualization environment, realize BigDataStack's vision of openness and extensibility.
BigDataStack delivers a complete high-performant stack of technologies addressing the needs of da... more BigDataStack delivers a complete high-performant stack of technologies addressing the needs of data operations and applications. The main objective of the dimensioning, modelling and interaction services building block of the BigDataStack environment, is to provide all the interaction mechanisms, including the Process Modelling framework, the Data Toolkit, the Dimensioning Workbench, and the Visualization environment. These are required in order to exploit the added-value services of the "underlying" BigDataStack offerings: the data-driven infrastructure management and the Data as a Service.
BigDataStack delivers a complete high-performant stack of technologies addressing the needs of da... more BigDataStack delivers a complete high-performant stack of technologies addressing the needs of data operations and applications. The main objective of the dimensioning, modelling and interaction services building block of the overall BigDataStack environment is to provide all the interaction mechanisms, including the Process Modelling framework, the Data Toolkit, the Dimensioning Workbench, and the Visualization environment. These are required in order to exploit the added-value services of the "underlying" BigDataStack data-driven infrastructure management and the Data as a Service offerings.
In the requirements analysis presented in this document, a top-down approach is taken with respec... more In the requirements analysis presented in this document, a top-down approach is taken with respect to the user requirements, which have been collected through the BigDataStack use case providers. This is complemented with a bottom-up approach aiming to identify, collect, and analyse the rest of the stakeholder requirements as well as technical requirements from the BigDataStack technology.
This deliverable presents Scientific Report and Prototype Description for the work carried out in... more This deliverable presents Scientific Report and Prototype Description for the work carried out in the first year of the BigDataStack project, related to the so-called Data-Driven Infrastructure Management capability of the BigDataStack platform. The document shows how the implementation of the solution is planned to be delivered following an incremental and iterative methodology, having cycles of implementation and experimentation. The document describes: the high-level assumptions and architecture of the capability, as well as detailed requirements, design and prototypes per component; the experimental use case scenarios and plans, as well as the experimental plan per component and its mapping with the use case scenarios.
This is the first version of the state-of-the-art and requirements analysis to drive the architec... more This is the first version of the state-of-the-art and requirements analysis to drive the architecture and research effort in BigDataStack. User requirements have been collected through the BigDataStack's use case providers and complemented with emerging technical requirements. They have also been tracked during the project lifetime so far to ensure that the BigDataStack platform will be fully addressed and properly considered.
The goal is to enable high scalability by decomposing a mixed-criticality application into a set ... more The goal is to enable high scalability by decomposing a mixed-criticality application into a set of "cloud-native" and "edge-running"<br> microservices, with different trust considerations, and managing secure accelerated offloading capabilities for distributing the resource intensive processes to the backend, thus, limiting the workload that needs to be managed at the edge. This will allow the overall system to reach its full potential, in a secure and trusted manner, without impeding safety.
The current document reflects the first deliverable of Innovation Management, which identifies an... more The current document reflects the first deliverable of Innovation Management, which identifies and specifies the approach of BigDataStack consortium for the Innovation. Moreover, it presents the priorities we set up based on Big Data Value Association (BDVA) priorities, the Innovation Strategy Plan and the IPR management approach. The first results of innovation monitoring throughout the different project activities are also presented. A section is devoted to each main building block of the architecture (i.e. data-driven infrastructure management, data-as-a-service, as well as dimensioning, modelling, and interaction services), presenting the innovation-monitored components, the related priorities, their expected innovation and the progress towards it. Another section describes the expected innovation of BigDataStack use cases and their approach. In the course of the project, the BigDataStack components and use cases will continue being monitored to ensure that the innovation potent...
This deliverable reports the exploitation activities carried out by BigDataStack partners during ... more This deliverable reports the exploitation activities carried out by BigDataStack partners during the second period of the project and the achievements in terms of exploitation and commercialization of the BigDataStack project. During the project lifetime, twenty-one exploitable assets have been identified, eight Minimum Valuable Products (MVP), as well as three exploitable business Use Cases. Also, three patents have arisen, and a technology transfer activity is in progress. Even though the COVID-19 pandemic situation, all exploitation activities have been properly performed. All the partners have developed their exploitation plans after the project, the partners have also already achieved remarkable exploitation results during the project, such as: commercialization of Data Skipping component by IBM and the Kuryr and Infrastructure API by RHT, Use Case partners are trying to leverage the use cases prototypes with their customers, most partners are already participating together in ...
The objective of this deliverable is to report the results of the activities performed in the fir... more The objective of this deliverable is to report the results of the activities performed in the first phase of CYRENE's Work Package 2. The main output is related to the requirements that have been collected from relevant standards and literature review, project pilot partners, as well as external stakeholders. The document can be divided in four main parts. In the first one, an overview of the Supply Chains is given, describing both their classification, including three different views (business, technical and sectorial) of the SCs, and their security aspects, consisting of the threat landscape, legal framework, SC security and Risk Management standards and SC risk assessment methodology and tools. In the second part, an overview of the EU Certification schemes is provided, encompassing the general definition and requirements (policy, legal, standards, methodologies, technical) regarding the security certification. Moreover, in the third part, the document reports on the methodol...
Cybele
HPC, Big Data, and AI Convergence Towards Exascale, 2022
Several visualization methods for eye tracking data exist to help researchers from many disciplin... more Several visualization methods for eye tracking data exist to help researchers from many disciplines depict data collected in eye tracking experiments. Focusing on eye tracking data from observations of cartographic lines, in this paper we propose a new visualization of eye tracking data using polylines inferred from the analysis of samples. This visualization depicts the "average" line that is actually seen by subjects; such a line can be useful in the study of various optical representation concepts, such as the assessment of the effects of alternative cartographic line attributes, distractions, abstraction levels and more, as well as in other cases such as the study of visual computer interfaces.
This is the Scientific Report and Prototype Description (Y2), reflecting the work done in the sco... more This is the Scientific Report and Prototype Description (Y2), reflecting the work done in the scope of the Data-Driven Infrastructure Management (DDIM) capability of the overall BigDataStack environment. The document describes the DDIM solution as assembled at M23 of the project (i.e. November 2019) in terms of updated design specifications, implementation, integration details, experimentation outcomes and next steps for the highlevel components comprising the DDIM solution: Cluster Management, Dynamic Orchestration, ADS Ranking & Deploy, Triple Monitoring & QoS Evaluation, and InformationDriven Networking. Regarding research results, it focuses on the research conducted to optimize the two components bringing artificial intelligence (AI) to the solution: the ADS Ranking—responsible for ranking and selecting the best application deployment configurations—and the Dynamic Orchestration—in charge of making re-deployment decisions. Both components make use of machine learning (ML) techn...
Map Construction from GPS Data
The use of virtual reality games, known as “exergaming”, is gaining more and more interest as a m... more The use of virtual reality games, known as “exergaming”, is gaining more and more interest as a mobilization tool and as a key piece in the delivery of quality health, especially in elderly people. Mobility tracking of elderly people facilitates the extraction of useful spatiotemporal characteristics regarding their activities and behavior at home. Currently, the analysis of human mobility is based on expensive technologies. In this paper, we propose a semantic interoperability agent which exploits mobility tracking and spatiotemporal characteristics to extract human profiling and give incentives for mobilization at home. The agent exploits an extended ontology which facilitates the collation of evidence for the effects of exergaming on the movement control of older adults. In order to provide personalized monitoring services, a number of rules are individually defined to generate incentives. To evaluate the proposed semantic interoperability agent, human mobility data are collected...
Uploads
Papers by Sophia Karagiorgou