Generative AI has been widely used in the higher education sector in a variety of ways. Some common activities include student assignments, but professors have also used them to prepare their classes, projects, and scientific articles. This article aims to conduct an exploratory study on the use of generative AI in different case studies related to higher education. The methodology used for this study is based on the framework for the use of generative AI proposed by UNESCO. Using this framework, we establish a set of criteria. Then, we analyze real-life case studies in different types of academic activities. From there, we identify the strengths and weaknesses of its use in each case study. We then identify the limitations and challenges of generative AI by type of academic activity. Finally, we conclude with recommendations on ethical responsibilities, and the role of the professor in the learning process, among other aspects.

Machine Learning algorithms are very sensitive to missing data, and their results depend on the way data is treated by the algorithms. In this work, we will understand missing data when a variable in a dataset is missing data but cannot be eliminated due to its relevance, or when some records of the dataset are missing the values of some of the variables. In general, these problems can be addressed during data preprocessing, for example, by removing variables with a lot of missing data. Now, our interest is to evaluate the robustness of the LAMDA (Learning Algorithm for Multivariable Data Analysis) algorithm to handle missing data at the variable level. A series of experiments will be carried out exposing the algorithm to cases with different percentages of missing variables, and the results obtained will be analyzed. In addition, the performance of the algorithm will be assessed compared to other machine learning algorithms in the context of classification and clustering problems. to carry out this study, several datasets are used. Multiple copies of each of them have been obtained, subjecting them to various transformations that allow us to study different case studies to test the effectiveness of the proposed algorithm.

In this work, an Automated Ontology Generator System (AOGS) is developed to autonomously create and populated ontologies based on the linked data paradigm. AOGS generates context-specific ontologies and populates these with information extracted from various linked data sources and pre-existing ontologies. AOGS allows reducing the time and effort required to create an ontology, ensuring consistency and accuracy in the ontology, and that domain experts to focus on higher-level tasks such as ontology refinement and evaluation. MEDAWEDE methodology is used to develop AOGS, which is then applied to create ontologies for COVID-19 and Energy. The quality of these ontologies is evaluated using precision, recall and f-measure metrics, which yield values between 0.9 and 0.7. The ontologies are also validated using Competence Questions, Consistency, and Quality Validation methods, all of which show good results.

In Machine Learning (ML), there is a lot of research that seek to automate specific processes carried out by data scientists in the generation of knowledge models (predictive, classification, clustering, etc.); however, an open problem is to find mechanisms that allow conferring the ability of self-learning. Thus, a meta-learning mechanism is required to allow ML techniques to self-adapt in order to improve their performance in problem solving, and even in some cases, to induce the learning algorithm itself. In this context, our research defines a meta-learning architecture using Linked Data (LD) for the automatic generation of knowledge models. Specifically, this intelligent architecture is formed by the layers of Knowledge Sources, Meta-Knowledge and Knowledge Modelling, to unify all processes to guarantee a Meta-Learning process. The Knowledge Sources layer is responsible for providing semantic knowledge about the processes of generation of knowledge models; the Meta-Knowledge layer is responsible for controlling the different processes and strategies for the automatic generation of knowledge models; and finally, the Knowledge Modelling layer is responsible for executing ML tasks defined by the Meta-Knowledge layer, among which are the tasks of feature engineering, ML algorithm configuration, model building, among others. Additionally, this article presents a case study to analyze the behavior of the different layers of the architecture, to generate knowledge models. Thus, the main contribution of this research is the definition of a Meta-Learning architecture for ML techniques, which takes advantage of the semantic information described as LD when generating the knowledge models. The preliminary results are very encouraging.

In this paper, we study the emergent behaviors in a multi-robot system. The multi-robot system uses a model for decision making that is composed of three levels: one individual, one collective, and another for the knowledge and learning management. In particular, the individual level, the base of the emergent behavior of the system, is composed of a module of perception/interpretation, an executing module and a behavioral module that has an emotional component, a reactive component, a cognitive component and a social component. In this paper, we analyze the robot performance, in order to produce an emergent behavior in the system. We present an example of an emergent scenario, and study its instantiation in our multi-robot architecture.