Conference Agenda

Cyberbullying in educational settings, fueled by the widespread use of social media, represents a growing threat to students’ emotional and academic well-being. In response, this study proposes a lightweight and explainable artificial intelligence model for the early detection of cyberbullying in digital comments. The objective was to design an automated, efficient, and interpretable system using the DistilBERT model within an MLOps framework, ensuring traceability, scalability, and continuous integration. The methodology included data collection from Twitter, text preprocessing, stratified supervised training, and evaluation using standard classification metrics. The results demonstrate that, when trained on 100% of the dataset, the model achieved a precision of 0.87, a recall of 0.83, and an average loss of 0.235—showing significant improvements over configurations using only 20% of the data. Qualitatively, the model successfully identified offensive language patterns with varying levels of subtlety and ambiguity. The integration of SHAP for explainability enabled real-time interpretation of predictions, enhancing the model’s transparency and trustworthiness. The study concludes that the proposed approach is suitable for implementation in schools and educational platforms, offering an accessible, interpretable, and effective tool for cyberbullying prevention. Future work is encouraged to extend this framework to multilingual models and multimodal analysis for broader applicability.

In the context of the highly competitive food sector in emerging regions, enhancing operational efficiency is essential for business sustainability. This study presents an innovative model aimed at improving the availability of oatmeal packaging machines in a Peruvian company by integrating emerging technologies with industrial engineering methodologies. The proposed approach combines the pillars of Autonomous and Planned Maintenance from the Total Productive Maintenance (TPM) framework with inventory management based on the SCOR model, incorporating a smart maintenance strategy through IoT sensors connected to a cloud-based platform. This technological solution enables real-time monitoring of machine conditions and supports data-driven operational decision-making. The implementation led to an 8.71% increase in machine availability (from 86.64% to 95.36%), a 68.68% reduction in unplanned downtimes, and a 78.78% decrease in stoppage time due to material and spare parts stockouts. Beyond the technical outcomes, this model represents a case of intrapreneurship within the operations team, demonstrating how internal innovation can transform traditional maintenance practices and contribute to the evolution of the business model. The solution is scalable and adaptable to other companies in the Latin American food industry, strengthening industrial competitiveness and promoting sustainable innovation in emerging markets.

The bottled water sector in Peru has shown sustained growth in recent years, driven by increased consumption and high seasonal demand, particularly during the summer. With a projected market value of $452.90 billion by 2029, companies in the industry face the ongoing challenge of adapting to a competitive and constantly changing environment. In this context, a bottling company identified improvement opportunities within its production system as it faced issues such as mechanical failures, supply shortages, planning errors, and rework, all of which compromised its operational efficiency. In response, strategies were promoted to optimize resource utilization, reduce unproductive time, and ensure product quality. These actions, driven from within the organization, reflected an intrapreneurial initiative focused on strengthening planning, preventive maintenance, and process standardization to enhance responsiveness to demand, particularly during peak months. Additionally, the company reinforced compliance with health and environmental regulations, contributing to the sustainability of its operations. As a result, the company achieved an 11% increase in operational efficiency, reduced losses, and enhanced its competitiveness. This experience illustrates how the pursuit of innovative solutions from within the work environment can transform traditional management practices, enabling organizations to adapt and thrive in a constantly evolving market while maintaining a commitment to efficiency, quality, and continuous improvement.

The incorporation of both Artificial Intelligence and automation in tax administration poses various challenges related to the impact on employment, the protection of taxpayers' rights and, in the same way, the need for policies that regulate their appropriate implementation. These tools are profoundly changing the tax systems of various countries, from automating processes such as fraud detection and the review of tax returns, as well as constant improvement in taxpayer service. This generates significant benefits such as increased efficiency, cost reduction and better tax compliance, which directly impacts the collection and economic development of the country.

In this sense, the objective of our research is to determine the perception that students of a university in Lima have about the implementation of Artificial Intelligence and automation in the Peruvian tax system, in the year 2025. On the other hand, the present work was based on a quantitative approach, and the design used was non-experimental, exploratory and cross-sectional, using the survey technique and through a series of questions asked in a questionnaire.

Thus, we were able to find that there is a high level of acceptance by the students who were surveyed on the topic raised who let us know their expectations and fears in this regard.

These young university students will play a key role in the modernization of the country. Although Peru has made progress in tax digitalization, there is still a gap in the integration of automated systems, which represents a technological and cultural challenge.

The rapid growth of decentralized blockchain-based financial systems has introduced new challenges in fraud detection, particularly due to anonymity, scalability, and the dynamic nature of transactions. Traditional supervised learning approaches often prove insufficient in these environments due to their reliance on labeled data and fixed patterns. This systematic literature review (SLR) investigates the application of unsupervised machine learning techniques for autonomous fraud detection in decentralized blockchain environments, analyzing their effectiveness, adaptability, and limitations compared to supervised or semi-supervised methods. Using a rigorous methodology based on the PICOC framework and PRISMA guidelines, we reviewed 20 studies published between 2021 and 2025. The results reveal that clustering and anomaly detection algorithms (such as autoencoders and graph-based methods) achieve superior performance (85–99% accuracy) in identifying common frauds like phishing and Ponzi schemes, leveraging blockchain's transparency and immutability. Key metrics such as recall (90–95%) and F1-score (88–93%) prove critical for evaluating models on imbalanced datasets. However, challenges persist in scalability, privacy, and cross-protocol adaptability. This study contributes a taxonomy of unsupervised techniques applied to blockchain fraud detection and proposes future research directions, such as hybrid models and standardized evaluation frameworks for decentralized ecosystems.

The use of simulation as a tool for emerging technologies allows logistics processes to be analyzed without directly intervening in reality. In this study, ProModel software was used to simulate and compare two inventory storage and dispatch strategies: one based on random allocation and the other using clustering techniques (cluster picking), as part of the Lean Logistics methodology. From a theoretical perspective, this methodology and tools such as simulation allow complex scenarios to be analyzed and improvements to be proposed without affecting actual operations. The research was carried out in five phases: problem analysis, data collection, modeling, simulation, and evaluation of results. For the clustering strategy, R Studio software was used, which allowed the optimal number of clusters to be defined and the inventory to be organized efficiently. Both models were evaluated based on the logistics process cycle time, resulting in a significant improvement in operational performance: cycle time, which was reduced by 25,30% with the implementation of the cluster-based strategy, search time was also reduced by 16,70%, and resource utilization increased by 38,90%. Overall, the simulated model based on cluster picking proves to be an effective alternative for reducing operating times and improving efficiency in distribution centers, contributing significantly to strategic logistics decision-making.

This paper analyzes the influence of artificial intelligence tools—specifically ChatGPT—on the academic performance of undergraduate students in the Software Engineering program at the Universidad Nacional Mayor de San Marcos. The study addresses a growing concern in higher education: the balance between leveraging AI for educational support and preserving the development of critical technical and cognitive skills. A mixed-methods approach was employed, combining quantitative surveys and qualitative interviews to assess how students use ChatGPT in relation to three key learning dimensions: conceptual understanding, practical skills, and problem-solving capabilities. Results indicate that while ChatGPT enhances task efficiency and offers real-time assistance, its overuse may negatively affect autonomous learning and deep understanding. Students recognize the tool's usefulness for explaining programming concepts and debugging code but also admit relying on it for tasks that demand analytical thinking. The study concludes that the responsible integration of generative AI in engineering education requires pedagogical strategies that promote critical reflection, ethical use, and alignment with long-term learning outcomes. The findings contribute to current debates on AI in education and offer evidence-based insights for faculty, curriculum designers, and policy makers.