Wheat is the most important and dominating crop in Pakistan in terms of production and acreage, which is grown on 37% of the cultivated area, accounting for 70% of the total production. However, wheat yield is highly affected by stripe rust, which is considered the most devastating fungal disease, causing 5.5 million tonnes of loss per year globally. In order to minimize this loss, the accurate and timely detection of rust disease is crucial instead of manual inspection. Towards this end, we propose a system to detect wheat rust disease and classify its infection types into four classes, including healthy, resistant, moderate (moderately resistant to moderately susceptible), and susceptible. The wheat rust dataset is collected indigenously from the National Agricultural Research Centre, Islamabad. A pre-trained U2 Net model is used to remove the background and extract the leaf containing the rust disease. Subsequently, two deep learning classifiers, including the Xception model and ResNet-50 are applied to classify the stripe rust severity levels, where the ResNet-50 model outperformed with the highest accuracy of 96%. This research presents a comparison between two state-of-the-art deep learning classifiers in terms of accuracy, memory utilization, and prediction time, which will assist the research community in selecting the most appropriate model for plant disease detection. Moreover, to assess the external validity, the performance of these classifiers is compared with the existing technique using a publicly available dataset, which confirms the validity of the results. Additionally, an intelligent edge computing rust detection device has been developed, where the trained ResNet-50 model is deployed, which facilitates the farmers to monitor the rust attack. The proposed research is aimed to assist the agricultural community to employ preventive measures in a site-specific manner based on the accurate diagnosis of rust disease & its severity, which is intended to improve the quality of the wheat as well as production.

This overview investigates the evolution and current landscape of eXplainable Artificial Intelligence (XAI) in healthcare, highlighting its implications for researchers, technology developers, and policymakers. Following the PRISMA protocol, we analyzed 89 publications from January 2000 to June 2024, spanning 19 medical domains, with a focus on Neurology and Cancer as the most studied areas. Various data types are reviewed, including tabular data, medical imaging, and clinical text, offering a comprehensive perspective on XAI applications. Key findings identify significant gaps, such as the limited availability of public datasets, suboptimal data preprocessing techniques, insufficient feature selection and engineering, and the limited utilization of multiple XAI methods. Additionally, the lack of standardized XAI evaluation metrics and practical obstacles in integrating XAI systems into clinical workflows are emphasized. We provide actionable recommendations, including the design of explainability‐centric models, the application of diverse and multiple XAI methods, and the fostering of interdisciplinary collaboration. These strategies aim to guide researchers in building robust AI models, assist technology developers in creating intuitive and user‐friendly AI tools, and inform policymakers in establishing effective regulations. Addressing these gaps will promote the development of transparent, reliable, and user‐centred AI systems in healthcare, ultimately improving decision‐making and patient outcomes.

Explainable AI (XAI) is increasingly used in clinical machine learning, yet quantitative evaluation of explanation quality is often reported inconsistently across methods and datasets. We present a reproducible, metric-driven framework for evaluating XAI methods on healthcare tabular data. The framework consolidates six established, family-specific metrics, fidelity, simplicity, consistency, robustness, precision, and coverage, into explicit equations; pairs them with a pre-specified focal-model protocol; and releases open-source code with a method–metric applicability map. We evaluate LIME, SHAP, Anchors, EBM, and TabNet across four public healthcare tabular datasets. Post-hoc explainers are applied to a single selected Random Forest focal predictor to control model-induced variability, whereas EBM and TabNet are assessed through their native interpretability mechanisms. Global explanation summaries are reported descriptively only. The results show that SHAP/TreeSHAP provides exact score reconstruction for the Random Forest setting, while LIME produces simpler but lower-fidelity explanations with greater instance-level variability. LIME and SHAP show the strongest rank agreement among the evaluated pairs, although agreement varies across datasets. TabNet often yields compact native explanations, but these must be interpreted alongside its dataset-specific predictive performance. EBM and TabNet show low sensitivity under the fixed Gaussian-jitter robustness protocol, while Anchors produces high-precision rules with reduced coverage at stricter thresholds. Overall, the framework enables controlled comparison under explicit method–metric and focal-model assumptions, supporting more transparent XAI selection for tabular machine learning. Although demonstrated in healthcare, the framework is transferable to other high-stakes tabular domains. Source code: https://github.com/matifq/XAI_Tab_Health .

The growing use of automated hate speech detection systems has sparked significant concern across technical, ethical, and sociopolitical dimensions. While recent advances in natural language processing and machine learning have improved classification performance, they often do so at the expense of transparency, fairness, and user trust. Explainable artificial intelligence (XAI) offers promising tools to bridge this gap, but current methods remain fragmented in scope and limited in stakeholder alignment. This review synthesizes the state of the art in explainability for hate speech detection by integrating machine learning, human‐computer interaction, and critical social science perspectives. We survey major XAI approaches, including ante‐hoc, post hoc, local, global, counterfactual, and rationale‐based methods, and evaluate their applicability across different stages of the ML pipeline and their relevance to key stakeholder groups such as developers, content moderators, policymakers, and affected communities. To guide this synthesis, we propose a conceptual framework that maps the intersection of explanation strategies, pipeline stages, and stakeholder needs. Using this model, we identify persistent gaps in dataset transparency, cultural and linguistic robustness, explanation evaluation practices, and participatory design processes. We argue that achieving meaningful explainability in content moderation goes beyond technical optimization; it demands sociotechnical alignment, contextual sensitivity, and accountability mechanisms that reflect the lived realities of those impacted by algorithmic decisions. The article concludes with interdisciplinary recommendations focused on dataset development, hybrid evaluation benchmarks, inclusive design, and integration. These strategies aim to foster hate speech detection systems that are not only more explainable but also more just, inclusive, and socially grounded. This article is categorized under: Commercial, Legal, and Ethical Issues > Social Considerations Fundamental Concepts of Data and Knowledge > Explainable AI Technologies > Machine Learning

The growing use of automated hate speech detection systems has sparked significant concern across technical, ethical, and sociopolitical dimensions. While recent advances in natural language processing and machine learning have improved classification performance, they often do so at the expense of transparency, fairness, and user trust. Explainable artificial intelligence (XAI) offers promising tools to bridge this gap, but current methods remain fragmented in scope and limited in stakeholder alignment. This review synthesizes the state of the art in explainability for hate speech detection by integrating perspectives from machine learning, human-computer interaction, and critical social science. We survey major XAI approaches, including antehoc, post-hoc, local, global, counterfactual, and rationale-based methods, and evaluate their applicability across different stages of the ML pipeline and their relevance to key stakeholder groups such as developers, content moderators, policymakers, and affected communities. To guide this synthesis, we propose a conceptual framework that maps the intersection of explanation strategies, pipeline stages, and stakeholder needs. Using this model, we identify persistent gaps in dataset transparency, cultural and linguistic robustness, explanation evaluation practices, and participatory design processes. We argue that achieving meaningful explainability in hate speech detection requires more than technical optimization; it demands sociotechnical alignment, contextual sensitivity, and accountability mechanisms that reflect the lived realities of those impacted by algorithmic decisions. The article concludes with interdisciplinary recommendations focused on dataset development, hybrid evaluation benchmarks, participatory interface design, and regulatory integration. Together, these strategies aim to foster hate speech detection systems that are not only more explainable but also more just, inclusive, and socially grounded.