4Experts: A Task Recommendation Approach Using Machine Learning and Biometric Data.

Context: The identification of skills and technical competences that a developer has and assigning tasks according to their abilities can improve task execution and reduce negative emotional aspects such as frustration. Problem: The assignment of coding tasks that require skills and technical competences greater than the competences of a developer can contribute to the expansion of the effort to perform a task and the generation of negative emotions such as frustration. Soluction:This study presents 4Experts, an approach to recommend software tasks, using machine learning and biometric data collected by an electroencephalogram (EEG), related to the brain activities of developers. IS Theory: This study was conceived under the aegis of the Cognitive Load Theory, related to the treatment of working memory by structuring information so that the individual can develop schemes and rules that allow storing learning in long-term memory, improving knowledge acquisition and performance. Method: The research has a prescriptive character, and its evaluation was carried out through proof of concept and controlled experiment. The results were analyzed using a quantitative approach. Summary of Results: In the experiments carried out, the 4Experts presented an accuracy of 83% in the prediction of the developer's expertise and an adherence to the use of the approach by 68% of the participants. Contributions and Impact in IS area: The main contribution of this approach is to recommend tasks through the use of biometric data that assess the developer's cognitive load, contributing to the acquisition of knowledge and performance in the construction of IS.