4157 papers
This study challenges the view of AI as a purely cognitive tool by demonstrating that, across 25 languages, AI-generated text systematically exhibits higher valence, arousal, and dominance with fewer negative emotions compared to human writing, revealing distinct affective profiles with significant implications for various applications.
This paper introduces MorphoSpan-Uz, a morphology-aware span-first model that enhances structured ABSA performance on Uzbek educational texts by integrating byte convolutions, suffix features, and boundary supervision, while establishing a rigorous evaluation protocol for cross-lingual Turkic transfer without claiming immediate cross-lingual gains.
This study demonstrates that an optimized EfficientNet-B0 model achieves a superior balance between facial emotion recognition performance and computational efficiency compared to larger EfficientNet variants, attaining 78.68% accuracy on a combined FER2013 and RAF-DB dataset while requiring significantly fewer parameters and less inference time.
INGREDIA is a hybrid recommendation system designed to suggest Indian recipes based on available ingredients, which outperforms a TF–IDF/KNN baseline in offline evaluations of ranking, coverage, and feasibility, though its real-world impact on user satisfaction and food waste remains unproven.
This study presents an explainable, lightweight DistilBERT-based system that improves online job offer screening by classifying postings into Real, Fake, and Invalid categories with 98.74% accuracy, addressing the limitations of traditional binary models through a balanced three-class approach and SHAP-based interpretability.
The paper proposes EViSE, an efficient multi-criteria summarization evaluation framework that combines cardinal-ordinal regression with within-source ranking to achieve high agreement with human judgments across faithfulness, coherence, and relevance while enabling fast, reference-free inference and effective use as a reward signal for policy optimization.
This paper introduces VGA-KIE, a modular post-hoc reliability framework that enhances Key Information Extraction from visually rich documents by coupling evidence-grounded state definition, typed failure diagnosis, and failure-conditioned repair with re-verification, achieving significant improvements in Field-F1 scores, evidence accuracy, and repair success rates across multiple benchmarks.
FedEAGD is a personalized federated learning framework that addresses data heterogeneity and partial client participation by employing a trajectory-guided, drift-aware distillation mechanism with adaptive historical knowledge transfer and progressive gated residual fusion to significantly improve model accuracy under non-IID conditions.
This paper presents MMIF, a GPU-accelerated parallel cryptographic permutation network that utilizes DNA-assisted chaotic permutation and XOR diffusion to securely and efficiently encrypt multimodal medical images for brain tumor detection, achieving high security metrics while enabling real-time processing.
This paper introduces FQXAI, a unified framework that enhances the interpretability and robustness of hybrid quantum-classical medical models by combining an exhaustive semantic explanation method (FQ-SHAP) with a noise-aware circuit-level analysis tool (SVQX) to generate human-readable rationales and optimize quantum gate configurations across diverse biomedical datasets.