Regresar

Selection of Rubrics using Fuzzy Algorithms

Abstract:

Within the teaching-learning process, evaluation involves various mechanisms generated by the teacher for its execution, such as identifying the elements to be evaluated, defining the process where it will take place, collecting information, and finally obtaining the assessment of the elements. The use of rubrics in the peer assessment process allows for assessing the achievements of students in the different proposed activities, but these must guarantee their validity and reliability to be considered as tools to support the teacher. Therefore, by implementing educational data mining techniques, it is possible to analyze the data from rubric validations, allowing for the selection of the most optimal ones for the teacher. This research presents the analysis of a clustering algorithm based on fuzzy logic C-means, which was constructed to facilitate decision-making in the appropriate selection of rubrics in Higher Education teaching-learning processes. To validate the algorithms, data generated in the laboratory with random values were used, the C-means algorithm was applied for k=4, verifying the internal validation of the clusters with the Silhouette Coefficient =0.282286 and the Dunn Coefficient =0.71947; furthermore, rubric selection was carried out using k=2 and k=4, yielding the same results in the selection of the rubrics, demonstrating that the developed algorithm using fuzzy logic is suitable in the rubric selection processes for peer assessment.