The Laboratory for Internet Computing (LInC) at the University of Cyprus received Best Paper Awards at three prestigious conferences during the Oct - Dec 2025: the 8th International Disaster and Risk Conference (IDRC-2025), the 16th IEEE International Conference on Knowledge Graph (ICKG-2025), and the 18th IEEE/ACM International Conference on Utility and Cloud Computing (UCC-2025).

IDRC 2025 Best Technical Innovation Paper

The paper titled "Next-Generation Emergency Communication Systems: Design and Evaluation of the e112 Mobile Application Platform", authored by Katerina loannidou, Marios D. Dikaiakos, and Athena Stassopoulou, was awarded the Best Technical Innovation Paper at IDRC 2025. Ms. Katerina loannidou presented the paper at the IDRC 2025 conference, which took place at the CYENS Centre of Excellence, Nicosia, Cyprus on October 22-24, 2025. The paper presents e112, a context-aware mobile emergency response application, designed to strengthen communication between citizens and authorities during disasters. Building on the ubiquity of smartphones, the system provides SOS requests, incident reporting, customized alerts, evacuation guidance, and moderated community interaction, supported by a cloud-based backend and an operator dashboard for situational awareness. A user-centered design approach guided the development of the system, ensuring clarity and usability under stressful conditions. Evaluation through usability studies and technical audits demonstrated high user satisfaction, robust performance, and accessibility. The results show that a simple, well-designed mobile application can significantly enhance emergency preparedness and response, reducing risks to human life during climate change–driven emergencies.

IEEE ICKG 2025 Best Paper Award

The paper titled "PRISM: A Framework for Multi-Level Modeling and Analysis of Polarization Knowledge", authored by Demetris Paschalides, George Pallis, and Marios D. Dikaiakos, was awarded the Best Paper Award at ICKG 2025. Dr. Demetris Paschalides presented the paper at the IEEE ICKG 2025 conference, which took place in Limassol, Cyprus, on November 13-14, 2025. The paper introduces a novel framework that models polarization using a typed, weighted, and directed structure known as the Polarization Knowledge Graph (PKG). PRISM provides analytical methods for multi-level analysis: (i) identifying key actors and categorizing them as protagonists or antagonists based on their contribution to conflict, (ii) measuring group cohesiveness through ideological alignment and topic-level agreement, and (iii) ranking topics by their polarization intensity. For validation, the authors present a case study on U.S. COVID-19 media discourse, uncovering polarization patterns that align with established findings and highlight the politicization of the pandemic.

IEEE/ACM UCC 2025 Best Paper Award

The paper titled "FakeInf: Selective Deep Neural Network Inference for Latency and Energy-Aware Model Serving Pipelines", authored by Demetris Trihinas, Moysis Symeonides, Nicolae Cleju, George Pallis, and Marios Dikaiakos, was awarded the Best Paper Award at UCC 2025. Dr. Moysis Symeonides presented the paper at the IEEE/ACM UCC 2025 conference, which took place in Nantes, France on December 01-04, 2025. The paper introduces a framework named FakeInf to support EdgeAI applications delivering DL-based video-stream inference. It is a lightweight framework that plugs into existing DL model-serving pipelines, continuously monitors the volatility of the analytics produced by the model, and decides, for each incoming stream datapoint, whether to run full DNN inference or selectively skip inference and "fake" outputs using low-cost statistical estimation, while keeping accuracy within user-defined QoS and imprecision limits. FakeInf was integrated and validated on a real smart-traffic system deployed on a MEC node, demonstrating substantial practical benefits: 59% lower application latency, 71% less network traffic, 66% lower GPU utilization, and 72% lower energy consumption, while incurring only a modest 4-6% reduction in the accuracy of emitted analytics. FakeInf also enabled the pipeline to process 2x more concurrent video streams before hitting inference bottlenecks.

More details can be found at https://linc.ucy.ac.cy