Background and objectives: Large language models (LLMs) show promise in automating open-ended evaluation tasks, yet their reliability in rubric-based assessment remains uncertain. Variability in scoring, feedback, and rubric adherence raises concerns about transparency and pedagogical validity in educational contexts. This study introduces CourseEvalAI, a framework designed to enhance consistency and fidelity in rubric-guided evaluation by fine-tuning a general-purpose LLM with authentic university-level instructional content. Methods: The framework employs supervised fine-tuning with Low-Rank Adaptation (LoRA) on rubric-annotated answers and explanations drawn from undergraduate computer science exams. Responses generated by both the base and fine-tuned models were independently evaluated by two human raters and two LLM judges, applying dual-layer rubrics for answers (technical or argumentative) and explanations. Inter-rater reliability was reported as intraclass correlation coefficient (ICC(2,1)), Krippendorff’s α, and quadratic-weighted Cohen’s κ (QWK), and statistical analyses included Welch’s t tests with Holm–Bonferroni correction, Hedges’ g with bootstrap confidence intervals, and Levene’s tests. All responses, scores, feedback, and metadata were stored in a Neo4j graph database for structured exploration. Results: The fine-tuned model consistently outperformed the base version across all rubric dimensions, achieving higher scores for both answers and explanations. After multiple-testing correction, only the Generative Pre-trained Transformer (GPT-4)—judged Technical Answer contrast remains statistically significant; other contrasts show positive trends without passing the adjusted threshold, and no additional significance is claimed for explanation-level results. Variance in scoring decreased, inter-model agreement increased, and evaluator feedback for fine-tuned outputs contained fewer vague or critical remarks, indicating stronger rubric alignment and greater pedagogical coherence. Inter-rater reliability analyses indicated moderate human–human agreement and weaker alignment of LLM judges to the human mean. Originality: CourseEvalAI integrates rubric-guided fine-tuning, dual-layer evaluation, and graph-based storage into a unified framework. This combination provides a replicable and interpretable methodology that enhances the consistency, transparency, and pedagogical value of LLM-based evaluators in higher education and beyond.

In the current context in which user authentication is the first line of defense against emerging attacks and can be considered a defining element of any security infrastructure, the need to adopt alternative, non-invasive, contactless, and scalable authentication mechanisms is mandatory. This paper presents initial research on the design, implementation, and evaluation of a multi-factor authentication mechanism that combines facial recognition with a fully homomorphic encryption algorithm. The goal is to minimize the risk of unauthorized access and uphold user confidentiality and integrity. The proposed device is implemented on the latest version of the Raspberry Pi and Arduino ESP 32 modules, which are wirelessly connected to the computer system. Additionally, a comprehensive evaluation, utilizing various statistical parameters, demonstrates the performance, the limitations of the encryption algorithms proposed to secure the biometric database, and also the security implications over the system resources. The research results illustrate that the Brakerski–Gentry–Vaikuntanathan algorithm can achieve higher performance and efficiency when compared to the Brakerski–Fan–Vercauteren algorithm, and proved to be the best alternative for the designed mechanism because it effectively enhances the level of security in computer systems, showing promise for deployment and seamless integration into real-world scenarios of network architectures.

This paper provides preliminary research regarding the implementation and evaluation of a hybrid mechanism of authentication based on fingerprint recognition interconnected with RFID technology, using Arduino modules, that can be deployed in different scenarios, including secret classified networks. To improve security, increase efficiency, and enhance convenience in the process of authentication, we perform a comparative assessment between two homomorphic encryption algorithms, the Paillier partial homomorphic algorithm and the Brakerski–Gentry–Vaikuntanathan fully homomorphic encryption scheme, applied to biometric templates extracted from the device mentioned above, by analyzing factors such as a histogram analysis, mean squared error (MSE), peak signal-to-noise ratio (PSNR), the structural similarity index measure (SSIM), the number of pixel change rate (NPCR), the unified average changing intensity (UACI), the correlation coefficient, and average encryption time and dimension. From security and privacy perspectives, the present findings suggest that the designed mechanism represents a reliable and low-cost authentication alternative that can facilitate secure access to computer systems and networks and minimize the risk of unauthorized access.

It is well known that bacterial plaque is the main etiological factor that causes the appearance of periodontal diseases and carious disease. Periodontal diseases can affect children and adolescents and are manifested in the form of gingivitis, but also the early form of chronic periodontitis as well as aggressive marginal periodontitis associated with local or general factors. Early periodontitis is frequently undiagnosed by clinicians due to a lack of knowledge of the specific symptoms. Certain systemic diseases, such as cardiovascular diseases, can create favorable conditions for the appearance and progression of severe manifestations of periodontal disease; also, recent research highlights that individuals with periodontal disease present an increased risk of developing cardiovascular diseases. Children with congenital or acquired cardiovascular diseases are at increased risk for complications resulting from the growth of microorganisms in the oral cavity, presenting a risk of infective endocarditis. The specific aim was to highlight the existing differences between the periodontal health of children with cardiovascular diseases and that of children without these diseases. The analyzed group included 124 patients, represented by children and adolescents, aged between 7 and 17 years, who were divided into four subgroups depending on the presence or absence of cardiovascular diseases and periodontal disease. A specialized clinical examination was performed for each patient, and periodontal clinical parameters were quantified (plaque index, gingival bleeding index, gingival index, community periodontal index of treatment needs) and associated with the diagnosis of general condition. Patients diagnosed with periodontal disease underwent specialized treatment and were called to a control visit 3 months after treatment. Statistical analysis showed significant differences between subgroups with much higher values of clinical parameters for patients with cardiovascular disease. Also, the response to the treatment was better in the case of patients in the control subgroup without cardiovascular diseases. The present study highlighted the interaction of three factors in the progression of periodontal diseases: subgingival microbiota, immune system response and environmental factors.

Large Language Models (LLMs) are increasingly used for rubric-based assessment, but reliable automated grading requires more than a single prompt-response step. This study presents GradeAgentOps, a verification-first framework for evidence-anchored LLM exam grading that combines strict grading contracts, deterministic verification and canonicalization, bounded semantic repair, optional memory modules, and provenance-aware logging. The framework is evaluated on a university-level dataset of 1000 short open-ended exam responses from 100 students across 10 questions, with annotations from two independent expert human graders. A controlled ablation protocol compares six configurations, including a rubric-only baseline and progressively stronger variants with repair and memory augmentation. Human–human agreement provides the reference context, with overall ICC(2,1) = 0.678 and QWK = 0.678 between the two graders. Using one expert grader as the operational reference, the FULL configuration achieves the strongest model-human agreement (MAE = 1.935, RMSE = 2.500, QWK = 0.652, Within ±2 = 0.667), with the consistency memory configuration with repair (C1) emerging as the closest alternative. The gains are not uniform: improvements are more pronounced on argumentative items than on technical ones, and pairwise comparisons show that consistency memory contributes more strongly than rubric memory alone. The stronger configurations also produce cleaner accepted outputs, with lower rates of evidence-related postprocess issues, although at a moderate operational cost. Overall, within this controlled single-course evaluation setting, the results show that reliable automated grading benefits not only from model capability, but also from pipeline design choices that promote verification, evidential coherence, and stable grading behavior.

Background and objectives: Automated evaluation of open-ended responses remains a persistent challenge, particularly when consistency, transparency, and reproducibility are required. While large language models (LLMs) have shown promise in rubric-based evaluation, their reliability across multiple evaluators is still uncertain. Variability in scoring, feedback, and rubric adherence raises concerns about interpretability and system robustness. This study introduces GraderAssist, a graph-based, rubric-guided, multi-LLM framework designed to ensure transparent and reproducible automated evaluation. Methods: GraderAssist evaluates a dataset of 220 responses to both technical and argumentative questions, collected from undergraduate computer science courses. Six open-source LLMs and GPT-4 (as expert reference) independently scored each response using two predefined rubrics. All outputs—including scores, feedback, and metadata—were parsed, validated, and stored in a Neo4j graph database, enabling structured querying, traceability, and longitudinal analysis. Results: Cross-model analysis revealed systematic differences in scoring behavior and feedback generation. Some models produced more generous evaluations, while others aligned closely with GPT-4. Semantic analysis using Sentence-BERT embeddings highlighted distinctive feedback styles and variable rubric adherence. Inter-model agreement was stronger for technical criteria but diverged substantially for argumentative tasks. Originality: GraderAssist integrates rubric-guided evaluation, multi-model comparison, and graph-based storage into a unified pipeline. By emphasizing reproducibility, transparency, and fine-grained analysis of evaluator behavior, it advances the design of interpretable automated evaluation systems with applications in education and beyond.