Abstract
As Artificial Intelligence (AI) systems increasingly permeate sensitive domains such as finance, healthcare, and media, ensuring their ethical deployment has become a central concern for researchers, policymakers, and practitioners. Current auditing tools often assess isolated principles, such as fairness or explainability, lacking a comprehensive view of the ethical risks involved. This paper presents a multidimensional framework for ethical evaluation of AI systems, designed to support responsible AI governance and alignment with the United Nations Sustainable Development Goals (SDGs). The proposed approach enables the simultaneous analysis of key ethical dimensions, including fairness, bias, explainability, robustness, transparency, and legal compliance. We demonstrate the applicability of this tool through one extensive case study: a credit scoring system, considered high-risk under the AI Act. This work contributes to operationalizing responsible AI governance, providing insight for policymakers, regulators, and practitioners to ensure ethical, legally compliant, and socially responsible AI deployment.

Abstract
Heating, Ventilation, and Air Conditioning (HVAC) systems represent a major share of building energy consumption, creating both operating cost and emissions reduction challenges for energy communities. This work proposes an optimisation-based management framework for a community of buildings, integrating HVAC operation with renewable generation, battery storage, and demand-side flexibility. The methodology employs a mixed-integer linear programming model to coordinate thermal and electrical energy flows, considering indoor comfort constraints, equipment dynamics, and market price signals. The framework is validated through a case study using real demand, weather, and market data, comparing baseline and optimised operation under varying seasonal conditions. Results demonstrate significant reductions in total operating cost and peak demand of the energy community, alongside improved usage of renewable generation and reduced reliance on the grid, without compromising thermal comfort. The proposed approach highlights the potential of coordinated HVAC scheduling in energy communities as a pathway toward more cost-efficient and sustainable building operation.

Abstract
Acoustic Event Detection and Classification (AEDC) aims to identify and classify specific audio events within audio signals. AEDC has applications in various fields, including security systems, scene monitoring, smart hospitals, environmental monitoring, and more. The process of AEDC typically involves steps that include audio signal processing to extract relevant features from the input, a machine learning model to recognise patterns in the extracted features and a classifier to detect events. Recent research on AEDC has increasingly focused on features based on the frequency distribution of the Mel-frequency cepstral coefficients (MFCCs). In this study, the feature extraction is performed based on Cochleogram, which involves the analysis of audio signals using Gammatone filters. Cochleogram features are inspired by the human cochlea, part of the inner ear responsible for converting sound vibrations into electrical signals sent to the brain. A two-dimensional (2D) feature is extracted from the Cochleogram using Welchs spectral density estimation and then converted into a frequency spectrum. The frequency distribution of different cochleogram filter banks is then used as a one-dimensional (1D) feature. The proposed classification method uses a 1D Convolutional Neural Network (CNN), which is less complex than traditional 2D CNNs. The proposed method was evaluated using the URBAN-SED dataset, and its performance was compared against the related state-of-the-art methods. The results showed the competitiveness of the cochleogram over Mel-based features such as MFCC in AEDC if the deep learning algorithm is properly designed and trained.

Abstract

Abstract
This study examines the role of online health communities in Brazil dedicated to cannabis treatments for chronic diseases as platforms for evidence-based activism. Using a mixed-methods approach, the research combines qualitative analysis with computational techniques, including Latent Dirichlet Allocation (LDA) topic modeling, to analyze six online groups from WhatsApp and Facebook. Key themes emerging from the analysis include treatment per pathology, treatment effects, access barriers, peer support, and advocacy efforts. The findings reveal how these communities act as epistemic networks, where patients and caregivers co-produce knowledge by sharing personal experiences and engaging in dialogue with healthcare professionals. This study highlights how online health communities transform experience sharing into structured evidence, enabling collective action to address barriers such as limited access to cannabis-based treatments. It underscores the potential of digital platforms to empower patients, foster collaboration with healthcare professionals, and influence health governance.

Abstract