3 May 2022 in Applied Intelligence
1 March 2026 in IEEE Transactions on Technology and Society
The increasing integration of artificial intelligence (AI) into everyday life makes it essential to explain AI-based decision-making in a way that is understandable to all users, including those with disabilities. Accessible explanations are crucial, as accessibility in technology promotes digital inclusion and allows everyone, regardless of their physical, sensory, or cognitive abilities, to use these technologies effectively. This paper presents a systematic literature review of the research on the accessibility of explainable artificial intelligence (XAI), specifically considering individuals with sight loss. Our methodology includes searching several academic databases with search terms to capture intersections between XAI and accessibility. The results of this survey highlight the lack of research on accessible XAI (AXAI) and stress the importance of including the disability community in XAI development to promote digital inclusion and accessibility and remove barriers. Most XAI techniques rely on visual explanations, such as heatmaps or graphs, which are inaccessible to individuals who are blind or have low vision. Therefore, it is necessary to develop explanation methods through non-visual modalities, such as auditory and tactile feedback; visual modalities accessible to persons with low vision; and personalized solutions that meet the needs of individuals, including those with multiple disabilities. We further emphasize the importance of integrating universal design principles into AI development practices to ensure that AI technologies are usable for everyone. We conclude the paper by discussing what constitutes a good explanation and desiderata for AXAI implementations.
31 March 2023 in IEEE Access
This paper proposes an edge-centric workload orchestration approach that uses machine learning in a three-tier vehicular architecture (edge, cloud via roadside unit, or cloud via cellular base station). The orchestrator uses a wireless network at the edge to receive and send over-the-air requests from vehicles, considering a metropolitan network connects the entire edge structure to support the high mobility of devices, allowing vehicles to share information and resources. Additionally, suppose the edge is congested, or its resources are unavailable. In that case, cloud resources will be used via roadside or cellular networks using a wide-area network to meet the tasks’ time constraints. Moreover, the proposed machine learning model uses variance-based sensitivity analysis to determine which inputs influence the model’s final decision. The experiments performed on the EdgeCloudSim simulator are based on modelling computational and network resources besides the representation of vehicles. The results indicate that our approach best fits task offloading over the air, outperforming the comparative experiments between the one-stage(our approach) model against two-stage and random models. Furthermore, by using our one-stage model that outputs the average of the prediction interval and the variance of this interval, we can measure how confident our model is in its prediction.
11 August 2023 in Research Square
Complex events originate from other primitive events combined according to defined patterns and rules. Instead of using specialists' manual work to compose the model rules, we use machine learning (ML) to self-define these patterns and regulations based on incoming input data to produce the desired complex event. Complex events processing (CEP) uncertainty is critical for embedded and safety-critical systems. This paper exemplifies how we can measure uncertainty for the perception and prediction of events, encompassing embedded systems that can also be critical to safety. Then, we propose an approach (ML\_CP) incorporating ML and sensitivity analysis that verifies how the output varies according to each input parameter. Furthermore, our model also measures the uncertainty associated with the predicted complex event. Therefore, we use conformal prediction to build prediction intervals, as the model itself has uncertainties, and the data has noise. Also, we tested our approach with classification (binary and multi-level) and regression problems test cases. Finally, we present and discuss our results, which are very promising within our field of research and work.