Journal of Data Engineering and Knowledge Discovery

Semantic Compatibility Auditing and Provenance Graph Construction for Heterogeneous Cybersecurity Features

Mon, 26 Oct 2026 00:00:00 +0000

Cross-dataset cybersecurity analysis can fail when similarly named variables represent different measurement procedures. This study performs a semantic compatibility audit using public mirrors of CICIoT2023 and CICIDS2017. The inspected IoT capture files contain 39 numerical fields, while the selected intrusion-flow derivative contains 78 numerical fields and two label columns. A compatibility ledger records 10 plausible feature correspondences, and a typed provenance graph separates dataset membership, capture provenance, and unresolved candidate matches. The resulting graph contains 132 nodes and 140 edges, with no asserted cross-source equivalence edge. A diagnostic classifier then attempts to identify the source dataset using seven superficially aligned numerical variables from benign records. With 8,000 training and 8,000 test observations, logistic regression achieves an ROC AUC of 0.9999995 and boosted trees achieve 1.0000. These values demonstrate strong source separability in the selected samples, but do not prove that any particular mapping is invalid. A first-principles aggregation counterexample explains why unit conversion alone cannot establish semantic equivalence. The completed study contributes an inspectable audit and provenance representation, while neither measuring improved attack transfer nor claiming discovery of a new cyber threat.

AI-Powered Personal Stylist: An Integrated Multimodal Architecture for Context-Aware Fashion Recommendation, Computer Vision Style Analysis, and Catalogue-Driven Outfit Composition

Syeda Alfiya Asif, Dinesh C — Tue, 27 Oct 2026 00:00:00 +0000

Online fashion retail offers consumers extensive product choices, yet assembling coordinated outfits that align with individual body characteristics, stylistic preferences, budget limits, and contextual occasions remains challenging. Unsuitable apparel selections contribute substantially to e-commerce return rates, which reached approximately 17.3% in retail reporting. This study presents the AI-Powered Personal Stylist, a modular web-based decision-support system designed to assist users in discovering cohesive fashion ensembles. The architecture couples a FastAPI backend and interactive frontend with a structured recommendation pipeline and a computer-vision-based image analysis module. Users provide explicit profile attributes, including sizing, preferred palettes, body silhouette, and financial boundaries, which are supplemented by an automated photo-analysis endpoint that extracts dominant clothing colors, garment patterns, lighting conditions, and aesthetic cues. To maintain pricing integrity and inventory feasibility, candidate garments are retrieved directly from a structured catalogue rather than generated through synthetic virtual try-on models. Furthermore, contextual parameters such as real-time weather and occasion-specific constraints are incorporated to refine outfit compositions. Uncertain visual estimates remain fully editable by the user, ensuring algorithmic transparency and personal agency. Functional testing demonstrates that integrating multi-attribute user profiles with catalogue-grounded filtering reliably produces balanced, four-piece coordinated looks with transparent styling rationale. The system provides an extensible foundation for transparent, context-aware digital fashion assistance while mitigating the operational and computational challenges associated with unconstrained generative styling.

Artificial Intelligence in Education: Impact on Students’ Critical Thinking, Logical Reasoning, and Cognitive Development a Comprehensive Review

Ashok Sadavare, Dattatraya Kumbhar, Aditya Sadavare — Wed, 26 Aug 2026 00:00:00 +0000

Educational institutions have adopted Artificial Intelligence (AI) tools at a pace that has outstripped careful study of their cognitive consequences for students. This paper synthesizes findings from 25 recent journal articles and empirical studies to examine how AI integration in school, undergraduate, and graduate settings shapes critical thinking, logical reasoning, and broader cognitive development. Generative large language models, intelligent tutoring systems, and adaptive learning platforms clearly strengthen personalized instruction and raise measurable academic performance; the evidence reviewed here, however, points to a separate risk: when such tools are adopted without safeguards, they can gradually erode independent reasoning, deductive problem-solving, and metacognitive self-regulation. One especially striking pattern emerges from the empirical literature: students who rely heavily on AI support tend to earn higher grades yet perform worse on assessments that require unassisted thinking, a divergence they term the grade-competence gap. To address this tension, the paper proposes the CLEAR Framework: contextual use, literacy, engagement, assessment reform, and reflection, a five-component approach for capturing AI's instructional value without undermining students' cognitive growth. The paper closes with practical recommendations for educators, curriculum designers, and policymakers navigating AI's growing role in education.