Information

This paper introduces ETQ-AI (Effective Teaching Questionnaire—Artificial Intelligence), a conversational AI framework designed to support teacher professional development by identifying pedagogical misconceptions and fostering critical reflection. Building on Evidence-Informed Education (EIE) principles, the system integrates a 50-item diagnostic questionnaire with an automated, Gemini-driven Explainable AI (XAI) feedback layer. When teachers complete the diagnostic form, they receive personalized feedback and can engage in dialogic interactions with Google Gemini to explore the underlying meta-analytic evidence (e.g., Hattie’s effect sizes, Cognitive Load Theory) or argue personal objections. We evaluated system fidelity against an expert panel benchmark ($N=64$ items) and conducted a pilot evaluation with 123 educators (70 primary teachers, 40 secondary teachers, 13 school leaders) recruited at an EIE conference. Gemini demonstrated high alignment with expert evaluations (97%), matching or exceeding static baseline feedback in user satisfaction (Mean = 4.1/5 vs. 3.8/5). Qualitative thematic analysis of 34 participant objections revealed key areas of tension, particularly around contextual constraints and pedagogical subjectivity. While the findings validate the technical proof-of-concept, system usability, and user engagement of ETQ-AI as an interactive scaffolding interface, we emphasize that this pilot study measures user perceptions and conversational patterns rather than direct pre/post shifts in critical reflection or classroom instructional practice.

Pest and disease detection in natural pear orchard environments is often affected by branch and leaf occlusion, non-uniform illumination, and complex background textures. In addition, substantial variations exist in the morphology and scale of lesions and insect-damaged regions, while the weak texture features of early-stage lesions can be progressively attenuated during feature propagation. Therefore, effectively representing targets at different scales while preserving discriminative local details and suppressing irrelevant background information is crucial for improving pest and disease detection in pear orchards. To address these challenges, this study proposes BAP-YOLO, a detection framework designed for natural pear orchard environments. The framework progressively enhances feature representation through context aggregation, adaptive scale modeling, and discriminative detail refinement. First, a Branch-based Target Enhancement Fusion (BTEF) module is introduced into the backbone network to enhance hierarchical feature representations at multiple backbone stages, thereby reducing interference from complex background information in target representation. Second, an Adaptive Scale Dynamic (ASD) module is designed and embedded into the neck network to construct candidate features with different effective receptive fields and dynamically modulate their contributions using multi-resolution statistics, thereby improving the model’s ability to detect targets with substantial scale variations. Finally, a Parallel Nonlinear Enhancement (PNE) module is incorporated between the neck network and detection head to enrich weak-texture and local-detail representations through complementary nonlinear transformations and element-wise fusion before prediction. Experiments were conducted on a pear orchard image dataset containing six classes: beetle, black_block, dry blight, health, rust, and scab. The experimental results showed that BAP-YOLO achieved a precision of 0.854, an mAP@50 of 0.876, and an F1-score of 0.826. Compared with the baseline YOLO11n, BAP-YOLO improved Precision, mAP@50, and F1-score by 6.1, 2.1, and 1.3 percentage points, respectively. These results demonstrate the potential of BAP-YOLO for pear disease and pest detection in natural orchard images.

Social media (SM) platforms are widely used for corporate communication, yet quantitative evidence from specialized agri-food sectors remains limited. This study evaluates the SM adoption, activity, audience growth, interaction efficiency, and interaction concentration of the 511 Greek enterprises registered for Protected Designation of Origin (PDO) and/or Protected Geographical Indication (PGI) olive oil. Adoption was modeled at the enterprise level using logistic regression, and key performance indicators for 116 Facebook, 70 Instagram, and 32 YouTube profiles were compared between the first halves of 2025 and 2026 using paired, multiplicity-corrected tests. Only 28.2% of enterprises maintained at least one official profile; adoption was higher among standardization- and commercialization-oriented enterprises and among companies than among sole proprietors. Cooperatives had the lowest descriptive adoption rate (13.1%). Follower counts increased significantly on all three platforms, but median gains were small, whereas no posting or interaction indicator changed significantly. Inactivity was persistent, inactive Facebook pages tended to lose followers, and interactions were highly concentrated, even among active profiles. Posting volume was associated with audience growth and cumulative interactions but not with interactions per post, and an exploratory analysis of message features found no association that was consistent across both periods. For certified producers, the key decision is less whether to open a profile than whether it can be sustained, and follower counts alone are a misleading measure of performance.