@article{Chaysri2023,
title = {Unmanned surface vehicle navigation through generative adversarial imitation learning},
author = {Piyabhum Chaysri, Christos Spatharis, Konstantinos Blekas, Kostas Vlachos},
url = {https://www.sciencedirect.com/science/article/pii/S0029801823013732},
doi = {https://doi.org/10.1016/j.oceaneng.2023.114989},
issn = {0029-8018},
year = {2023},
date = {2023-06-13},
journal = {Ocean Engineering},
volume = {282},
pages = {114989},
abstract = {In the artificial intelligent and big data technology era, the marine industry among others is inevitably developing in this direction, aiming at becoming autonomous and completing tasks without relying on human involvement while providing safety. The technology of small unmanned surface vehicles (USVs) is relatively mature but with a large development potential and wide research interest expecting significant benefits such as safety and high efficiency in shipping and transportation systems. This article addresses these issues and utilizes an imitation learning algorithm to resolve autonomous navigation for USVs even in complex environmental conditions. We formulate the trajectory modeling as a data-driven imitation learning problem where we employ a state of the art imitation learning algorithm. Experiments are performed in a particular simulated environment tailored to match the specific weather conditions of the local area. The simulation results show the potential of the proposed imitation learning scheme to create advanced intelligent agents for USVs under real-world environmental settings, and USV actuation constraints that allow to predict trajectories with high accuracy and safety.
In addition, we evaluated the method’s robustness in generating successful trajectories under environmental conditions that differed from those encountered during training, thereby promoting knowledge reusing without the need for retraining.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Bastas2023,
title = {Data-Driven Modeling of Air Traffic Controllers’ Policy to Resolve Conflicts},
author = {Alevizos Bastas, George A. Vouros},
doi = {https://doi.org/10.3390/aerospace10060557},
issn = {2226-4310},
year = {2023},
date = {2023-06-13},
journal = {Aerospace},
volume = {10},
number = {6},
abstract = {With the aim to enhance automation in conflict detection and resolution (CD&R) tasks in the air traffic management (ATM) domain, this article studies the use of artificial intelligence and machine learning (AI/ML) methods to learn air traffic controllers’ (ATCOs) policy in resolving conflicts among aircraft assessed to violate separation minimum constraints during the en route phase of flights, in the tactical phase of operations. The objective is to model ℎ𝑜𝑤 conflicts are being resolved by ATCOs. Towards this goal, the article formulates the ATCO policy learning problem for conflict resolution, addresses the challenging issue of an inherent lack of information in real-world data, and presents AI/ML methods that learn models of ATCOs’ behavior. The methods are evaluated using real-world datasets. The results show that AI/ML methods can achieve good accuracy on predicting ATCOs’ actions given specific conflicts, revealing the preferences of ATCOs for resolution actions in specific circumstances. However, the high accuracy of predictions is hindered by real-world data-inherent limitations.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Bastas2023b,
title = {Data-Driven Modeling of Air Traffic Controllers’ Policy to Resolve Conflicts},
author = {Alevizos Bastas, George A Vouros},
doi = {https://doi.org/10.3390/aerospace10060557},
year = {2023},
date = {2023-06-13},
journal = {Aerospace},
volume = {10},
number = {6},
pages = {557},
abstract = {With the aim to enhance automation in conflict detection and resolution (CD&R) tasks in the air traffic management (ATM) domain, this article studies the use of artificial intelligence and machine learning (AI/ML) methods to learn air traffic controllers’ (ATCOs) policy in resolving conflicts among aircraft assessed to violate separation minimum constraints during the en route phase of flights, in the tactical phase of operations. The objective is to model how conflicts are being resolved by ATCOs. Towards this goal, the article formulates the ATCO policy learning problem for conflict resolution, addresses the challenging issue of an inherent lack of information in real-world data, and presents AI/ML methods that learn models of ATCOs’ behavior. The methods are evaluated using real-world datasets. The results show that AI/ML methods can achieve good accuracy on predicting ATCOs’ actions given specific conflicts, revealing the preferences of ATCOs for resolution actions in specific circumstances. However, the high accuracy of predictions is hindered by real-world data-inherent limitations.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Koletis2023,
title = {Can Semantics Uncover Hidden Relations between Neurodegenerative Diseases and Artistic Behaviors?},
author = {Adam Koletis, Pavlos Bitilis, Nikolaos Zafeiropoulos, Konstantinos Kotis},
url = {https://www.mdpi.com/2076-3417/13/7/4287},
doi = {https://doi.org/10.3390/app13074287},
year = {2023},
date = {2023-03-28},
journal = {Applied Sciences},
volume = {13},
number = {7},
pages = {4287},
abstract = {Semantics play a crucial role in organizing domain knowledge, schematizing it, and modeling it into classes of objects and relationships between them. Knowledge graphs (KGs) use semantic models to integrate and represent different types of data. This study aimed to systematically review related work on the topics of ontologies for neurodegenerative diseases (NDs), ontology-based expert systems for NDs, and the artistic behavior of ND patients. The utilization of ontologies allows for a more comprehensive understanding of the progression and etiology of NDs, the structure and function of the brain, and the artistic expression associated with these diseases. The data collected from ND patients highlights the presence of cases where artistic expression can be linked to the disease. By developing fuzzy ontologies for NDs and incorporating them into expert systems, early detection and monitoring can be supported. Through our systematic review, we identify and discuss open issues and challenges in understanding the relationship between ND patients and their artistic behavior. We also conclude that ontology-based expert systems hold immense potential in uncovering hidden correlations between these two. Further research in this area has the potential to address key research questions and provide deeper insights.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Pliatsios2023,
title = {A systematic review on semantic interoperability in the IoE-enabled smart cities},
author = {Antonios Pliatsios, Konstantinos Kotis, Christos Goumopoulos},
url = {https://www.sciencedirect.com/science/article/pii/S254266052300077X},
doi = {https://doi.org/10.1016/j.iot.2023.100754},
issn = {2542-6605},
year = {2023},
date = {2023-03-22},
journal = {Internet of Things},
volume = {22},
pages = {100754},
abstract = {Smart cities have emerged as a result of smart interconnections of people, processes, data, and things, representing an excellent case study of the Internet of Everything (IoE) paradigm. One of the main challenges in realizing the smart city vision is how to provide seamless interoperability between the IoE entities. In this paper we conduct a systematic literature review on the use of semantic technologies to support interoperability between IoE entities in smart cities, with the goal of identifying the main trends and challenges in adopting semantic interoperability solutions for sustainable, green, and resilient smart cities. To this end, we have extracted data from selected primary studies over the last decade that address semantic interoperability issues in smart cities through related technologies and techniques such as ontologies, linked open data, knowledge graphs, ontology alignment/matching methods, and automated reasoning mechanisms. We have analyzed the maturity of this research area by exploring three research questions that focus on: i) the importance of semantic interoperability in the smart city domain; ii) the identification of semantic technologies and tools applied in the smart city domain to promote semantic interoperability; and iii) the identification of smart city application areas where semantic technologies are used to efficiently deliver smart services. The analysis provided research insights, including the introduction of a new evaluation framework that assesses semantic interoperability solutions on four maturity levels. The framework includes specific evaluation criteria for attributes such as modeling, scalability, and availability. Finally, an elaborated list of strengths, opportunities, weaknesses, and threats of semantic interoperability solutions in smart cities is provided, along with a discussion of open challenges and future work in this domain.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Karaliopoulos2023,
title = {Beyond clustering: Rethinking the segmentation of energy consumers when nudging them towards energy-saving behavior},
author = {Merkouris Karaliopoulos, Leonidas Tsolas, Iordanis Koutsopoulos, Maria Halkidi, Stephanie Van Hove, Peter Conradie},
url = {https://dl.acm.org/doi/abs/10.1145/3584024.3584028},
doi = {https://doi.org/10.1145/3584024.3584028},
year = {2023},
date = {2023-02-13},
journal = {ACM SIGEnergy Energy Informatics Review},
volume = {2},
number = {4},
pages = {28-43},
abstract = {Besides technological innovations in energy production and management technologies, the fight against climate change requires fundamental changes in our energy consumption behavior. Behavioral interventions are key to this process, especially when tailored to different energy consumer segments accounting for their socio-demographic profiles, socio- psychological characteristics and energy consumption practices.
In this work, we propose a novel approach to energy consumer segmentation that facilitates the choice of (nudging) interventions for each segment. We call it intervention-driven energy consumer profiling since it explicitly considers upfront the set of interventions that can be delivered to energy consumers and defines profiles that can be readily matched with them. The profiles are specified as combinations of socio-psychological factors with implications for energy-saving behavior and are parameterized by thresholds that measure how strongly these factors are represented in each profile. One profile represents ideal energy-savers, whereas each of the remaining five profiles shares one or two distinct features that serve as barriers towards energy-saving behavior and/or prescribe specific type of nudging interventions for strengthening such behavior. We use the responses of users to a European-wide online survey to formulate and solve an optimization problem for these thresholds and then assign the survey respondents to the profiles. Finally, we analyze them also in terms of socio-demographic variables and recommend appropriate nudging interventions for them.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Koryzis2023,
title = {Disruptive technologies for parliaments: A literature review},
author = {Dimitris Koryzis, Dionisis Margaris, Costas Vassilakis, Konstantinos Kotis, Dimitris Spiliotopoulos},
url = {https://www.mdpi.com/1999-5903/15/2/66/pdf?version=1676550074},
doi = {https://doi.org/10.3390/fi15020066},
year = {2023},
date = {2023-02-05},
journal = {Future Internet},
volume = {15},
number = {2},
pages = {66},
abstract = {Exploitation and use of disruptive technologies, such as the Internet of Things, recommender systems, and artificial intelligence, with an ambidextrous balance, are a challenge, nowadays. Users of the technologies, and stakeholders, could be part of a new organisational model that affects business procedures and processes. Additionally, the use of inclusive participatory organisational models is essential for the effective adoption of these technologies. Such models aim to transform organisational structures, as well. Public organisations, such as the parliament, could utilise information systems’ personalisation techniques. As there are a lot of efforts to define the framework, the methodology, the techniques, the platforms, and the suitable models for digital technologies adoption in public organisations, this paper aims to provide a literature review for disruptive technology inclusive use in parliaments. The review emphasises the assessment of the applicability of the technologies, their maturity and usefulness, user acceptance, their performance, and their correlation to the adoption of relevant innovative, inclusive organisational models. It is argued that the efficient digital transformation of democratic institutions, such as parliaments, with the use of advanced e-governance tools and disruptive technologies, requires strategic approaches for adoption, acceptance, and inclusive service adaptation.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Kravaris2022,
title = {Explaining deep reinforcement learning decisions in complex multiagent settings: towards enabling automation in air traffic flow management},
author = {Theocharis Kravaris, Konstantinos Lentzos, Georgios Santipantakis, George A Vouros, Gennady Andrienko, Natalia Andrienko, Ian Crook, Jose Manuel Cordero Garcia, Enrique Iglesias Martinez},
url = {https://pmc.ncbi.nlm.nih.gov/articles/PMC9169601/pdf/10489_2022_Article_3605.pdf},
doi = {https://doi.org/10.1007/s10489-022-03605-1},
year = {2023},
date = {2023-02-01},
journal = {Applied Intelligence},
volume = {53},
number = {4},
pages = {4063-4098},
abstract = {With the objective to enhance human performance and maximize engagement during the performance of tasks, we aim to advance automation for decision making in complex and large-scale multi-agent settings. Towards these goals, this paper presents a deep multi agent reinforcement learning method for resolving demand – capacity imbalances in real-world Air Traffic Management settings with thousands of agents. Agents comprising the system are able to jointly decide on the measures to be applied to resolve imbalances, while they provide explanations on their decisions: This information is rendered and explored via appropriate visual analytics tools. The paper presents how major challenges of scalability and complexity are addressed, and provides results from evaluation tests that show the abilities of models to provide high-quality solutions and high-fidelity explanations.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Kotis2023b,
title = {A kg-based integrated uav approach for engineering semantic trajectories in the cultural heritage documentation domain},
author = {Konstantinos Kotis, Sotiris Angelis, Efthymia Moraitou, Vasilis Kopsachilis, Ermioni-Eirini Papadopoulou, Nikolaos Soulakellis, Michail Vaitis},
url = {https://www.mdpi.com/2072-4292/15/3/821},
doi = {https://doi.org/10.3390/rs15030821},
year = {2023},
date = {2023-01-31},
journal = {Remote Sensing},
volume = {15},
number = {3},
pages = {821},
abstract = {Data recordings of the movement of vehicles can be enriched with heterogeneous and multimodal data beyond latitude, longitude, and timestamp and enhanced with complementary segmentations, constituting a semantic trajectory. Semantic Web (SW) technologies have been extensively used for the semantic integration of heterogeneous and multimodal movement-related data, and for the effective modeling of semantic trajectories, in several domains. In this paper, we present an integrated solution for the engineering of cultural heritage semantic trajectories generated from unmanned aerial vehicles (UAVs) and represented as knowledge graphs (KGs). Particularly, this work is motivated by, and evaluated based on, the application domain of UAV missions for documenting regions/points of cultural heritage interest. In this context, this research work extends our previous work on UAV semantic trajectories, contributing (a) an updated methodology for the engineering of semantic trajectories as KGs (STaKG), (b) an implemented toolset for the management of KG-based semantic trajectories, (c) a refined ontology for the representation of knowledge related to UAV semantic trajectories and to cultural heritage documentation, and (d) the application and evaluation of the proposed methodology, the developed toolset, and the ontology within the domain of UAV-based cultural heritage documentation. The evaluation of the integrated UAV solution was achieved by exploiting real datasets collected during three UAV missions to document sites of cultural interest in Lesvos, Greece, i.e., the UNESCO-protected petrified forest of Lesvos Petrified Forest/Geopark, the village of Vrissa, and University Hill.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Kravaris2023,
title = {Explaining deep reinforcement learning decisions in complex multiagent settings: towards enabling automation in air traffic flow management},
author = {Theocharis Kravaris, Konstantinos Lentzos, Georgios Santipantakis, George A. Vouros, Gennady L. Andrienko, Natalia V. Andrienko, Ian Crook, Jose Manuel Cordero Garcia, Enrique Iglesias Martinez},
doi = {10.1007/s10489-022-03605-1},
year = {2023},
date = {2023-01-20},
journal = {Applied Intelligence},
volume = {53},
pages = {4063-4098},
abstract = {With the objective to enhance human performance and maximize engagement during the performance of tasks, we aim to advance automation for decision making in complex and large-scale multi-agent settings. Towards these goals, this paper presents a deep multi agent reinforcement learning method for resolving demand – capacity imbalances in real-world Air Traffic Management settings with thousands of agents. Agents comprising the system are able to jointly decide on the measures to be applied to resolve imbalances, while they provide explanations on their decisions: This information is rendered and explored via appropriate visual analytics tools. The paper presents how major challenges of scalability and complexity are addressed, and provides results from evaluation tests that show the abilities of models to provide high-quality solutions and high-fidelity explanations.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Mulita2023,
title = {Deep Learning for Colon Cancer: Our Experience and a Review of the Literature},
author = {Francesk Mulita, Georgios-Ioannis Verras, Konstantinos Kotis, Christos-Nikolaos Anagnostopoulos},
url = {https://www.sciencedirect.com/science/article/abs/pii/S0748798322008150},
doi = {https://doi.org/10.1016/j.ejso.2022.11.087},
issn = {0748-7983},
year = {2023},
date = {2023-01-17},
journal = {European Journal of Surgical Oncology},
volume = {49},
number = {1},
pages = {e15-e16},
abstract = {Methods: We systematically searched PubMed from inception to 30 August 2022 for primary studies developing a DL model for the histopathological interpretation of large intestine biopsy tissues and CRC. The search was conducted on 4 September 2022.
Results: Our systematic search returned 201 articles, 106 of which were selected for full-text screening. Finally, 96 articles were considered eligible for our systematic review according to our criteria of eligibility. A detailed description of the study selection process can be found in the PRISMA flowchart.
Conclusions: When dealing with human disease, particularly cancer, we need in our armamentarium all available resources, and DL applications have recently become very popular in medical image analysis due to the effects and successes it has achieved in the early detection and screening of cancerous tissue or organ.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Kontogiannis2023b,
title = {Xdqn: Inherently interpretable dqn through mimicking},
author = {Andreas Kontogiannis, George Vouros},
doi = {https://doi.org/10.48550/arXiv.2301.03043},
year = {2023},
date = {2023-01-08},
journal = {arXiv},
abstract = {Although deep reinforcement learning (DRL) methods have been successfully applied in challenging tasks, their application in real-world operational settings is challenged by methods’ limited ability to provide explanations. Among the paradigms for explainability in DRL is the interpretable box design paradigm, where interpretable models substitute inner constituent models of the DRL method, thus making the DRL method “inherently” interpretable. In this paper we explore this paradigm and we propose XDQN, an explainable variation of DQN, which uses an interpretable policy model trained through mimicking. XDQN is challenged in a complex, real-world operational multi-agent problem, where agents are independent learners solving congestion problems. Specifically, XDQN is evaluated in three MARL scenarios, pertaining to the demand-capacity balancing problem of air traffic management. XDQN achieves performance similar to that of DQN, while its abilities to provide global models’ interpretations and interpretations of local decisions are demonstrated.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Kotis2023,
title = {ReconTraj4Drones: A Framework for the Reconstruction and Semantic Modeling of UAVs’ Trajectories on MovingPandas},
author = {Konstantinos Kotis, Andreas Soularidis},
url = {https://www.mdpi.com/2076-3417/13/1/670},
doi = {https://doi.org/10.3390/app13010670},
issn = {2076-3417},
year = {2023},
date = {2023-01-03},
journal = {Applied Sciences},
volume = {13},
number = {1},
pages = {670},
abstract = {Unmanned aerial vehicles (UAVs), also known as drones, are important for several application domains, such as the military, agriculture, cultural heritage documentation, surveillance, and the delivery of goods/products/services. A drone’s trajectory can be enriched with external and heterogeneous data beyond latitude, longitude, and timestamp to create its semantic trajectory, providing meaningful and contextual information on its movement data, enabling decision makers to acquire meaningful and enriched contextual information about the current situation in the field of its operation and eventually supporting simulations and predictions of high-level critical events. In this paper, we present an ontology-based, tool-supported framework for the reconstruction, modeling, and enrichment of drones’ semantic trajectories. This framework extends MovingPandas, a widely used and open-source trajectory analytics and visualization tool. The presented research extends our preliminary work on drones’ semantic trajectories by contributing (a) an updated methodology for the reconstruction of drones’ trajectories from geo-tagged photos taken by drones during their flights in cases in which flight plans and/or real-time movement data have been lost or corrupted; (b) an enrichment of the reconstructed trajectories with external data; (c) the semantic annotation of the enriched trajectories based on a related ontology; and (d) the use of SPARQL queries to analyze and retrieve knowledge related to the flight of a drone and the field of operations (context). An evaluation of the presented framework, namely, ReconTraj4Drones, was conducted against several criteria, using real and open datasets.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Spatharis2023,
title = {Hierarchical multiagent reinforcement learning schemes for air traffic management},
author = {Christos Spatharis, Alevizos Bastas, Theocharis Kravaris, Konstantinos Blekas, George A. Vouros, Jose Manuel Cordero Garcia},
doi = {10.1007/s00521-021-05748-7},
year = {2023},
date = {2023-01-01},
journal = {Neural Computing and Applications},
volume = {35},
pages = {147-159},
abstract = {In this work we investigate the use of hierarchical multiagent reinforcement learning methods for the computation of policies to resolve congestion problems in the air traffic management domain. To address cases where the demand of airspace use exceeds capacity, we consider agents representing flights, who need to decide on ground delays at the pre-tactical stage of operations, towards executing their trajectories while adhering to airspace capacity constraints. Hierarchical reinforcement learning manages to handle real-world problems with high complexity, by partitioning the task into hierarchies of states and/or actions. This provides an efficient way of exploring the state–action space and constructing an advantageous decision-making mechanism. We first establish a general framework of hierarchical multiagent reinforcement learning, and then, we further formulate four alternative schemes of abstractions, on states, actions, or both. To quantitatively assess the quality of solutions of the proposed approaches and show the potential of the hierarchical methods in resolving the demand–capacity balance problem, we provide experimental results on real-world evaluation cases, where we measure the average delay per flight and the number of flights with delays.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Vouros2023,
title = {Explainable Deep Reinforcement Learning: State of the Art and Challenges},
author = {George A. Vouros},
doi = {10.1145/3527448},
year = {2023},
date = {2023-01-01},
journal = {ACM Computing Surveys},
volume = {55},
number = {5},
pages = {1-39},
abstract = {Interpretability, explainability, and transparency are key issues to introducing artificial intelligence methods in many critical domains. This is important due to ethical concerns and trust issues strongly connected to reliability, robustness, auditability, and fairness, and has important consequences toward keeping the human in the loop in high levels of automation, especially in critical cases for decision making, where both (human and the machine) play important roles. Although the research community has given much attention to explainability of closed (or black) prediction boxes, there are tremendous needs for explainability of closed-box methods that support agents to act autonomously in the real world. Reinforcement learning methods, and especially their deep versions, are such closed-box methods. In this article, we aim to provide a review of state-of-the-art methods for explainable deep reinforcement learning methods, taking also into account the needs of human operators—that is, of those who make the actual and critical decisions in solving real-world problems. We provide a formal specification of the deep reinforcement learning explainability problems, and we identify the necessary components of a general explainable reinforcement learning framework. Based on these, we provide a comprehensive review of state-of-the-art methods, categorizing them into classes according to the paradigm they follow, the interpretable models they use, and the surface representation of explanations provided. The article concludes by identifying open questions and important challenges.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}

@article{Spatharis2023b,
title = {Hierarchical multiagent reinforcement learning schemes for air traffic management},
author = {Christos Spatharis, Alevizos Bastas, Theocharis Kravaris, Konstantinos Blekas, George A Vouros, Jose Manuel Cordero},
url = {https://link.springer.com/article/10.1007/s00521-021-05748-7},
doi = {https://doi.org/10.1007/s00521-021-05748-7},
year = {2023},
date = {2023-01-01},
journal = {Neural Computing and Applications},
volume = {35},
number = {1},
pages = {147-159},
abstract = {In this work we investigate the use of hierarchical multiagent reinforcement learning methods for the computation of policies to resolve congestion problems in the air traffic management domain. To address cases where the demand of airspace use exceeds capacity, we consider agents representing flights, who need to decide on ground delays at the pre-tactical stage of operations, towards executing their trajectories while adhering to airspace capacity constraints. Hierarchical reinforcement learning manages to handle real-world problems with high complexity, by partitioning the task into hierarchies of states and/or actions. This provides an efficient way of exploring the state–action space and constructing an advantageous decision-making mechanism. We first establish a general framework of hierarchical multiagent reinforcement learning, and then, we further formulate four alternative schemes of abstractions, on states, actions, or both. To quantitatively assess the quality of solutions of the proposed approaches and show the potential of the hierarchical methods in resolving the demand–capacity balance problem, we provide experimental results on real-world evaluation cases, where we measure the average delay per flight and the number of flights with delays.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}