Home » Actividades (Page 5)
Category Archives: Actividades
¡Que viene la IA! ¿estoy preparada/o?
¡Que viene la IA! ¿estoy preparada/o?
David Fonseca. La Salle. Universidad Ramon Llull.
Francisco José García Peñalvo. Universidad de Salamanca.
Faraón Llorens. Universidad de Alicante.
Rafael Molina. Universidad de Alicante.
Conferencia colaborativa e interactiva sobre Inteligencia Artificial
CINAIC 2023 (Congreso Internacional sobre Aprendizaje, Innovación y Cooperación)
http://cinaic.net
Presentación:
En zenodo: https://zenodo.org/records/10050857
(pdf)
Resumen: pdf
Acceso al video de la conferencia:
Presentación utilizada: pdf
Incrementando la madurez digital en sistemas universitarios iberoamericanos
Panel
Incrementando la madurez digital en sistemas universitarios iberoamericanos
Ponentes: Dr. Faraón Llorens, Dr. Antonio Fernández y Dra. Alicia Daverio.
Moderadoras: Dra. Luz María Castañeda de León, Dra. Alejandra Herrera y Dra. Carmen Díaz
9 de octubre de 2023
https://encuentro-tic.anuies.mx/panel-madurez-digital
Encuentro ANUIES-TIC 2023
“La transformación digital de la educación superior para una nueva sociedad”
9, 11, 12 y 13 de octubre de 2023
Universidad Autónoma de San Luis Potosí
Observatorio de Inteligencia Artificial en Educación Superior
Hoy a tenido lugar en Madrid en la UE STEAM School (Escuela de Arquitectura, Ingeniería y Diseño) la primera reunión presencial del Observatorio de Inteligencia Artificial en Educación Superior de la Universidad Europea.
Muchas gracias por contar conmigo. Ha sido una sesión muy interesante y en la que he aprendido mucho.
La Universidad Europea constituye un observatorio de Inteligencia Artificial
Decoding Student Error in Programming: An Iterative Approach to Understanding Mental Models
Decoding Student Error in Programming: An Iterative Approach to Understanding Mental Models
Francisco J. Gallego-Durán, Patricia Compañ-Rosique, Carlos J. Villagrá-Arnedo, Gala M. García-Sánchez, Rosana Satorre-Cuerda, Rafael Molina-Carmona, Faraón Llorens-Largo, Sergio J. Viudes-Carbonell, Alberto Real-Fernández & Jorge Valor-Lucena
25th International Conference on Human-Computer Interaction (HCII 2023)
Copenhagen, Denmark
23-28 July 2023
10th International Conference on Learning and Collaboration Technologies (LCT 2023).
Lecture Notes in Computer Science book series (LNCS,volume 14040) (https://link.springer.com/book/10.1007/978-3-031-34411-4)
https://link.springer.com/chapter/10.1007/978-3-031-34411-4_18
https://doi.org/10.1007/978-3-031-34411-4_18
Abstract
In computer programming education, despite yearly changes in teaching methodologies, students still struggle to grasp the concepts. When they advance to more complex projects, gaps in their basic knowledge become evident. It seems that the knowledge they learn in the first course is forgotten or not well understood. This proposal aims to explore students’ mental models of computer programming concepts to better understand and identify any misconceptions. An iterative methodology is proposed to identify, test, analyse and evidence students’ erroneous mental models in programming. Characterising these mental models is a first step to deepen our understanding and designing strategies to help students improve them. The proposed methodology is exemplified in detail through an undergoing use case at the University of Alicante, and some early results are discussed.
Keywords
Programming, Learning, Mental Models
Cite this paper as:
Gallego-Durán, F.J. et al. (2023). Decoding Student Error in Programming: An Iterative Approach to Understanding Mental Models. In: Zaphiris, P., Ioannou, A. (eds) Learning and Collaboration Technologies. HCII 2023. Lecture Notes in Computer Science, vol 14040. Springer, Cham. https://doi.org/10.1007/978-3-031-34411-4_18
Two-Phases AI Model for a Smart Learning System
Two-Phases AI Model for a Smart Learning System
Javier García-Sigüenza, Alberto Real-Fernández, Rafael Molina-Carmona & Faraón Llorens-Largo
25th International Conference on Human-Computer Interaction (HCII 2023)
Copenhagen, Denmark
23-28 July 2023
10th International Conference on Learning and Collaboration Technologies (LCT 2023).
Lecture Notes in Computer Science book series (LNCS,volume 14040) (https://link.springer.com/book/10.1007/978-3-031-34411-4)
https://link.springer.com/chapter/10.1007/978-3-031-34411-4_4
https://doi.org/10.1007/978-3-031-34411-4_4
Abstract
Current Information Technologies are mature enough to favor the creation of adaptive learning systems that also encourages active, autonomous and persistent learning. A solution could be the creation of artificial intelligence algorithms capable of detecting the individual learning needs and features of the learners, what skills they are acquiring and how they do it, or how they behave, in order to offer them an adapted and personalized learning experience. This is what is defined a smart learning system.
Therefore, in this research we aim to propose an Artificial Intelligence (AI) model for a learning system to achieve this purpose. It is based on a learning model called CALM (Customized Adaptive Learning Model), that offers personalized learning through different learning paths and adapts to each learner by offering a specific activity at any time. The selection of this activity relies on an AI engine that detects the needs and characteristics of the learner and selects the most appropriate activity.
To implement an AI model for this purpose, applying CALM principles, we propose the use of both the information provided by activities and the learner’s characteristics and progression. Combining these datasets with the use of deep learning techniques, we propose a two phases process. First, the model makes predictions that are personalized for each student, and then it applies a concrete instructional strategy to make the final decision, allowing the teacher to adapt and guide the student’s learning.
Keywords
Smart Learning, Artificial Intelligence, Deep Learning
Cite this paper as:
García-Sigüenza, J., Real-Fernández, A., Molina-Carmona, R., Llorens-Largo, F. (2023). Two-Phases AI Model for a Smart Learning System. In: Zaphiris, P., Ioannou, A. (eds) Learning and Collaboration Technologies. HCII 2023. Lecture Notes in Computer Science, vol 14040. Springer, Cham. https://doi.org/10.1007/978-3-031-34411-4_4
Educando centauros digitales
Educando centauros digitales
Faraón Llorens
Revista Hipótesis
https://www.ull.es/portal/cienciaull/revistahipotesis/
Nº 15. Más allá del ChatGPT. Reflexiones sobre el presente y futuro de la IA
https://www.ull.es/portal/cienciaull/tabletplanet/?w=4806
Universidad de La Laguna
¿Estamos enseñando a nuestros jóvenes lo que saben hacer las máquinas? Esto sería un error imperdonable, ya que las máquinas lo harán no solo mejor, sino más barato y sin cansarse ni pedir vacaciones. Hemos de aprender a colaborar con las máquinas, no a luchar contra ellas, ya que el resultado de esta colaboración será superior al que se consigue de forma separada.
/…/
Soy optimista, lo sé, pero va en el ADN de ser profesor. Tenemos un sistema educativo diseñado para un mundo con escasez de información, en el que había que ir a buscarla y guardarla para cuando la necesitásemos. Eso justificaba la etapa de nuestras vidas en la que nos formábamos y acudíamos a la universidad. Pero ahora vivimos en una sociedad con sobreinformación (verdades, medias verdades y falsedades), con acceso inmediato y a demanda a la misma. Las universidades, cumpliendo nuestro compromiso con la sociedad de creación, transmisión y preservación del conocimiento, ¿sabremos dar respuesta a este reto?
Usos y desusos del modelo GPT-3 entre estudiantes de grados de ingeniería
Usos y desusos del modelo GPT-3 entre estudiantes de grados de ingeniería
Daniel Amo-Filva, David Fonseca, David Vernet, Eduard De Torres, Pol Muñoz Pastor, Víctor Caballero, Eduard Fernandez, Marc Alier Forment, Francisco José García-Peñalvo, Alicia García-Holgado, Faraón Llorens-Largo, Rafael Molina-Carmona, Miguel Á. Conde, Ángel Hernández-García
XXIX Jornadas sobre Enseñanza Universitaria de la Informática (JENUI 2023)
Granada, 5, 6 y 7 de julio de 2023
Resumen
La herramienta ChatGPT, basada en el modelo GPT-3 desarrollado por OpenAI, ya se utiliza por estudiantes de grados de ingeniería como herramienta de apoyo en su proceso de aprendizaje. En este contexto, las implicaciones negativas que presenta el uso de esta herramienta son diversas: dependencia tecnológica, obstaculización del saber y conocer práctico, error en las respuestas, problemas éticos o incluso problemas legales. El uso de esta herramienta sin que los estudiantes hayan recibido formación se considera como problema a investigar. El objetivo es entender en profundidad el contexto tecnológico de la herramienta, cómo se utiliza actualmente entre los estudiantes de ingeniería de un conjunto de universidades privadas y públicas, y su impacto en la educación universitaria. Este artículo es un trabajo en desarrollo donde se presenta el contexto del estudio, la metodología de investigación y unos primeros resultados. Se conduce una encuesta cualitativa-exploratoria con una muestra de más de 360 estudiantes de grados de ingeniería matriculados en diferentes cursos. Se utiliza una estratificación aleatoria para asegurar que la muestra sea representativa de la población. Los resultados sugieren que el modelo GPT-3 puede ser utilizado como una herramienta beneficiosa para los estudiantes de grados de ingeniería.
Palabras clave
ChatGPT, GPT-3, OpenAI, TIC, universidad, ingeniería, riesgos tecnológicos, proceso de aprendizaje, ética, legalidad, encuesta cualitativa-exploratoria, estratificación aleatoria.
Artículo completo (pdf)
Vol 8 (2023) – Actas de las XXIX Jornadas sobre Enseñanza Universitaria de la Informática (Granada, 5, 6 y 7 de julio de 2023) (pdf)
Few-Shot Learning for Prediction of Electricity Consumption Patterns
Few-Shot Learning for Prediction of Electricity Consumption Patterns
Javier García-Sigüenza, José F. Vicent, Faraón Llorens-Largo y José-Vicente Berná-Martínez
IbPRIA 2023: 11th Iberian Conference on Pattern Recognition and Image Analysis
Alicante, Spain. June 27-30, 2023
www.ibpria.org/2023
Publicación:
Pertusa, A., Gallego, A.J., Sánchez, J.A., Domingues, I. (eds) Pattern Recognition and Image Analysis. IbPRIA 2023. Lecture Notes in Computer Science, vol 14062. Springer, Cham.
https://doi.org/10.1007/978-3-031-36616-1_56
Abstract
Deep learning models have achieved extensive popularity due to their capability for providing an end-to-end solution. But, these models require training a massive amount of data, which is a challenging issue and not always enough data is available. In order to get around this problem, a few shot learning methods emerged with the aim to achieve a level of prediction based only on a small number of data. This paper proposes a few-shot learning approach that can successfully learn and predict the electricity consumption combining both the use of temporal and spatial data. Furthermore, to use all the available information, both spatial and temporal, models that combine the use of Recurrent Neural Networks and Graph Neural Networks have been used. Finally, with the objective of validate the approach, some experiments using electricity data of consumption of thirty-six buildings of the University of Alicante have been conducted.
Keywords
Few-shot learning, Graph neural networks, Electricity consumption, Pattern recognition
Metamodelo para implementação de transformação digital em IES
Metamodelo para implementação de transformação digital em IES:
jornada de transformação por meio de abordagem multiteórica de mudança organizacional
Adriana Veríssimo Karam Koleski
Programa De Pós-Graduação Em Engenharia E Gestão Do Conhecimento
Centro Tecnológico
Universidade Federal De Santa Catarina
Florianópolis, 2023
Hoy he formado parte de la banca examinadora de esta tesis doctoral.
Parabéns, Adriana!
ABSTRACT
The transformations experienced in the context of the networked society and the impact of digital technologies in all areas of society, have brought challenges and opportunities to organizations and, in particular, to higher education institutions (HEIs). The challenge of HEIs is twofold: at the same time that they have to transform as organizations to remain competitive, they need to reconfigure the education offered to their students. A process of planning, implementation and monitoring of strategies that enable their digital transformation (DT) is necessary. A transformation that is nourished by scientific knowledge and technologies structured by society and, at the same time, has a unique character for each HEI. This thesis addresses the question of how to implement DT in HEIs. It was was structured with the objective of conceiving a metamodel for DT implementation that respects the diversity, complexity and scope of the phenomenon in a HEI. The challenge was addressed using organizational change theories accompanied by concepts about DT in higher education, networked society, stakeholder theory and 21st century competencies. The nature of the research is technological and the paradigm adopted was pragmatism. As for the objectives it is exploratory and prescriptive, with the use of the mixed method and Design Science Research (DSR) as its methodological approach. The metamodel was conceived using design cycles sustained in the state of the art literature on the subject. The artifact is composed of four elements: digital transformation journey; theoretical lenses of change and digital transformation factors; stakeholders involved; spiraling journey. The results of evaluation of the metamodel were obtained by means of structured and semi-structured interviews with domain experts and strategic managers of HEIs and reveal that the metamodel is consistent, feasible and useful to guide the construction of DT models for HEIs.
Keywords: digital transformation; higher education; organizational change; metamodel.
Explainability techniques applied to road traffic forecasting using Graph Neural Network models
Explainability techniques applied to road traffic forecasting using Graph Neural Network models
Javier García-Sigüenza, Faraón Llorens-Largo, Leandro Tortosa and José F. Vicent
Information Sciences
Volume 645, October 2023, 119320
doi: doi.org/10.1016/j.ins.2023.119320
Available online 16 June 2023
(INS 119320)
https://www.sciencedirect.com/science/article/pii/S0020025523009052
Abstract
In recent years, several new Artificial Intelligence methods have been developed to make models more explainable and interpretable. The techniques essentially deal with the implementation of transparency and traceability of black box machine learning methods. Black box refers to the inability to explain why the model turns the input into the output, which may be problematic in some fields. To overcome this problem, our approach provides a comprehensive combination of predictive and explainability techniques. Firstly, we compared statistical regression, classic machine learning and deep learning models, reaching the conclusion that models based on deep learning exhibit greater accuracy. Of the great variety of deep learning models, the best predictive model in spatio-temporal traffic datasets was found to be the Adaptive Graph Convolutional Recurrent Network. Regarding the explainability technique, GraphMask shows a notably higher fidelity metric than other methods. The integration of both techniques was tested by means of experimental results, concluding that our approach improves deep learning model accuracy, making such models more transparent and interpretable. It allows us to discard up to 95% of the nodes used, facilitating an analysis of its behavior and thus improving the understanding of the model.
Keywords: Graph neural networks, deep learning, data analysis, explainability, traffic flow