Webavalia: Algorithm Improvement of an eAssessment Tool Based on a User Survey

There has been an increased use of workgroups during Problem-Based Learning (PBL) activities, which have implied the emergence of self and peer evaluation practices in education. In turn, these practices have highlighted the importance of e-assessment tools where the collection of feedback is regarded. The consideration of the students’ self and peer assessment about a project has influence on their individual mark. Literature states that e-assessment tools need to consider restrictions in the students’ distribution of scores during the self and peer assessment. An emerging technology that has recently started to be applied in the development of software tools is Machine Learning. This technology allows the possibility to implement algorithms that can process the students’ feedback about their marks, increasing the efficiency on the methodology implemented. This paper presents a framework that collects feedback of a survey and translates those results in parameter values that will dynamically change the distribution function of the final individuals’ mark. The advantages and restrictions of different solutions for the framework, based on the literature, are also discussed. Finally, there is the presentation of WEBAVALIA, a free software tool where the framework will be implemented, improving its usability, user experience, flexibility, and efficiency.