Conferencia
Permanent URI for this collectionhttps://hdl.handle.net/11285/636053
Presentación o disertación realizada dentro de un congreso o evento similar, o como evento académico independiente, tales como: Conferencia inaugural, conferencia magistral, conferencia de clausura.
Browse
Search Results
- Design and challenges of open large language model frameworks (Open LLM): a systematic literature mapping(IATED Digital Library, 2024-11-11) García López, Iván Miguel; Ramírez Montoya, María Soledad; Molina Espinosa, José Martín; https://ror.org/03ayjn504Analyzing the frameworks of open large language models (OLLM) is essential to understanding how the management of these artificial intelligence (AI) models can be regulated. This study aims to analyze the evidence published from 2019 to 2024 regarding OLLM frameworks that integrate AI. Systematic mapping was the method for reviewing 227 articles published in the Scopus and Web of Science (WoS) databases. Inclusion, exclusion, and quality criteria filtered the papers to obtain the maximum relevant information. The analysis and classification of articles related to open LLM frameworks and models yielded significant findings per our research questions. The challenges identified were a) improving customization and accuracy through open LLMs, b) latency and efficiency challenges, c) the importance of reliability and security, and d) complex operational management (LLMOps). This review provides a framework for identifying the topic's state of the art and current and emerging research trends.
- Generative artificial intelligence in higher education: a literature mapping perspective(IATED Digital Library, 2024-11-11) García López, Iván Miguel; Ramírez Montoya, María Soledad; Molina Espinosa, José Martín; https://ror.org/03ayjn504Generative Artificial Intelligence (GAI) opens intriguing possibilities for creating unique and innovative content. This systematic literature review (SLR) analysed the published evidence on GAI in higher education institutions from January 2018 to October 2023. Fourteen articles on the topic were identified in the Web of Science (WOS) and Scopus databases. We screened to obtain the most relevant data by applying inclusion, exclusion, and quality criteria. The results identified (1) the characteristics of the publications, (2) methodological trends, and (3) the approaches implemented. The study aims to be of value to the academic community and developers of GAI initiatives
- Regulatory challenges and optimization strategies for open large language models: a multidimensional framework for efficient management(Springer Link, 2024-10-23) García López, Iván Miguel; Ramírez Montoya, María Soledad; Molina Espinosa, José Martín; https://ror.org/03ayjn504Innovations in artificial intelligence are rapidly transforming various in-dustries, particularly through the development and deployment of Open Large Language Models (OLLMs). However, the absence of a robust regula-tory framework presents significant challenges in ensuring the ethical, safe, and effective use of these models. This research aims to address this gap by proposing a comprehensive regulatory framework designed to optimize the scalability and performance of OLLMs, emphasizing the importance of structured pruning techniques. By integrating both quantitative and qualita-tive analyses, the study will assess the technical capabilities and societal implications of OLLMs, ultimately providing clear guidelines that promote responsible and sustainable innovation. The expected outcomes include the development of a governance model that balances the need for innovation with ethical considerations, offering a pathway for the regulation of OLLMs that supports their continued evolution while safeguarding public interests.
- Modelos de lenguaje grande, la revolución silenciosa en educación, salud e industria: mapeo sistémico de literatura(International Insitute of Systemics, Cybernetics, and Informatics: IIIS, 2024-09-10) García López, Iván Miguel; Ramírez Montoya, María Soledad; Molina Espinosa, José Martín; https://ror.org/03ayjn504Los Modelos de Lenguaje Grande (LLMs) están revolucionando múltiples sectores mediante su capacidad para procesar y generar texto con un alto nivel de coherencia y contexto. Estos modelos no permiten acceso público a su código fuente ni datos de entrenamiento, lo que restringe su uso a la organización que los posee. A diferencia de los modelos de código abierto, los LLMs cerrados presentan tanto oportunidades como desafíos únicos. Nos preguntamos: ¿Cuáles son las estrategias efectivas para integrar LLMs en educación, salud e industria, y cómo pueden manejar los retos éticos, de seguridad y transparencia? El método del estudio es una revisión sistemática de la literatura, analizando artículos de las bases de datos Scopus y Web of Science desde enero de 2019 hasta mayo de 2024. Se seleccionaron los datos más relevantes utilizando criterios de inclusión, exclusión y calidad, y se delimitaron 60 artículos para su análisis. Los hallazgos destacan: (a) oportunidades significativas en la personalización y eficiencia en diversos sectores, (b) marcos de trabajo para la integración efectiva de LLMs y (c) estrategias para abordar desafíos éticos y de seguridad. Este escrito invita a otros investigadores a explorar el uso de LLMs en distintos ámbitos, resaltando su potencial para transformar procesos educativos, médicos e industriales, mientras se asegura un uso responsable y ético.
- Unveiling success: an analysis of academic performance predictors in a private high school in mexico through learning analytics(Springer Link, 2024-08-05) Velarde Camaqui, Davis; Peláez Sánchez, Iris Cristina; Viehmann, Cristina Mariana; https://ror.org/03ayjn504The article explores the effect of fathers’ educational level and high school type on the academic performance of students in three faculties of a private university in Mexico. Learning analytics helps to this research, so the study uses a quantitative analysis and a two-factor ANOVA model to examine the relationship between socioeconomic variables and academic variables on student performance. The dataset includes sociodemographic and academic information of first-semester students, as well as enrollment in different academic programs. The results show that students who graduated from the same system preparatory school have higher academic performance compared to students who did not. Additionally, the educational level of the father has a significant effect on academic performance. When analyzing results by academic career, it is observed that the effect of type of high school on academic performance is significant in the Medicine population but not in the Engineering population. There is also a significant difference in academic performance between the type of high school graduates and non-graduates in the Engineering population. In conclusion, the study highlights the importance of socioeconomic variables and high school characteristics in predicting academic performance. The findings can provide insights for improving educational practices and interventions to enhance student outcomes.
- Social robotics and development of investigative skills: complex thinking as a pivotal notion to produce academic documents(Springer Link, 2024-08-05) George Reyes, Carlos Enrique; López Caudana, Edgar Omar; Suárez Brito, Paloma; https://ror.org/03ayjn504To establish a relationship between social robotics (SR) and the development of research skills (HI) in university students, a hinge notion must be found that allows both terms to overlap and from there lay the foundations that favor the construction of teaching strategies. Disruptive initiatives based on educational innovation and mediated by the use of humanoid robots (NAO) that serve to scale basic academic research skills. The purpose of this study is to propose a referential scaffolding that enables the scaling of investigative skills in higher education students through the imbrication of the components of complex thinking (CP) with the characteristics and dimensions of social robotics. The result is a framework that proposes to understand the way in which the RS-PC-HI relationship can be used to positively influence the performance of students to appropriately elaborate products related to scientific writing. It is concluded that by using the proposed frame of reference it will be possible to generate formative and motivating experiences that generate not only knowledge related to academic research but also its derived products.
- Scaling social, scientific, and technological entrepreneurship skills: designing the OpenEdR4C platform(Springer Link, 2024-08-05) George Reyes, Carlos Enrique; López Caudana, Edgar Omar; Suárez Brito, Paloma; Alcantar Nieblas, Carolina; https://ror.org/03ayjn504The development of entrepreneurship skills in higher education is necessary to train citizens to identify opportunities and solve problems in an increasingly complex society. However, training at the university is not enough, so experiences must be designed that help students strengthen social, scientific and technological entrepreneurship skills through a hinge competence: complex thinking. This paper shows a proposal for the design of a web platform through a project called OpenEdR4C that imbricates complex thinking-entrepreneurship and whose purpose is to contribute to the training of students and lifelong learners with skills and competencies necessary to face the challenges of today’s world and develop their potential as entrepreneurs and committed and innovative citizens. The description of the OpenEdR4C project, its objectives, stages, as well as the first pre-prototype that will become a learning platform is presented.
- Simulations for learning in complex scenarios: students’ most valued elements(Springer Link, 2024-08-05) Pacheco Velázquez, Ernesto Armando; Rodés Paragarino, Virginia; https://ror.org/03ayjn504Game-based learning is an effective approach to developing learning skills, and simulations play a crucial role in the logistics field by providing realistic and hands-on training. This article presents the findings of a four-year study that investigated students’ experiences and perceptions of a Logistics Simulator, known as LOST, used in a logistics undergraduate course for engineering education in Mexico. LOST simulated supply chain operations and aimed to enhance the development of logistical concepts and complex skills. Qualitative data was collected through open-ended questions to assess students’ experiences with LOST, involving a total of 216 students across different course cohorts. The responses were analyzed using axial coding to identify the most valued elements of the game-based learning experience with the logistics simulator. The results indicate that students highly valued elements such as the Real World, Decision-Making, Theory in Practice, Learning, Knowledge, Complexity, Experimentation, and Strategies. These findings emphasize the transformative potential of game-based experiences with the simulator, offering innovative and engaging learning opportunities for educational communities, equipping students with crucial decision-making skills, and providing decision-makers with insights into the effectiveness of simulation-based approaches for enhancing learning outcomes and preparing individuals for real-world complex challenges.
- Lifelong learning and education 4.0 to strengthen the attraction of extramural research funding(Springer Link, 2024-08-05) Miranda Mendoza, Jhonattan; Kreiner Baumgartner, Isabel; Alvarez-Icaza Longoria, Inés; https://ror.org/03ayjn504This work presents the role of Education 4.0 in lifelong learning to strengthen the attraction of extramural research funding. This approach recognizes the need for continuous learning in academia and advocates for the lifelong learning of researchers during their mature careers. This work also addresses the challenges identified around lifelong learning for researchers and the existing challenges in training to attract extramural research funding to researchers. Also, it is shown how, by applying the enablers of Education 4.0, it is possible to enable smart learning environments (SLE), including products, processes, and infrastructure, to support the attraction of extramural research funding. Finally, the case study “Writing Successful Grant Proposal Lifelong Learning Program for Researchers” illustrates how new lifelong learning and tools strengthen the attraction of extramural research funding are emerging to strengthen the attraction of extramural research funding.
- Equity in hyperconnectivity: are there gender differences in the metaverses?(Springer Link, 2024-08-05) Peláez Sánchez, Iris Cristina; Velarde Camaqui, Davis; https://ror.org/03ayjn504Digital natives have the opportunity to live in a hyperconnected world; however, differences still exist in society, such as the gender digital divide. Thus, the metaverse represents an option for hyperconnected interaction, although there is still a knowledge gap regarding the differences in interaction among young people in this digital environment. The study’s main objective was to compare the interaction of digital natives in the metaverse. The study population (n = 88) was established based on the age of adolescents between 15 and 29 years old identified as digital natives with constant digital interaction. The study results demonstrate that metaverses are virtual environments where young people can interact similarly, as evidenced by the Mann-Whitney U test (p = .769 > .05). However, each dimension showed notable results. For instance, regarding behavior, women (n = 19, 47.5%) exhibited positive behavior, while men considered their behavior hostile within the metaverse (n = 18, 37.5%).