The student performance has been affected for different factors, many of them are unobvious. The habits or daily activities undoubtly have a deep effects on the student performance. In this work, the study of student daily activities, and the relationship with his academic performance, using Data Mining techniques was done. In the attribute selection phase, 5-13 attributes from the 35 total were selected. The students were classified in four classes related with their academic performance: low, regular, good and high; the classification accuracy was near to 90%, using algorithms like MLP, KNN and tree algorithms like Random Forest, Random Tree and J48. The activities and factors presented for low and high performance students, also the tendency of activities and factors in the four classes, are reported.

The teaching-learning processes for algebra are crucial, because they promote the cognitive bases for the study of other knowledge areas like engineering's; however, its topics are one of the most complicated to understand for many students. In the last years, the virtual learning environments have been taking importance to support the teaching-learning process in general way. In this work the development of a based games virtual learning environment, containing elementals topics of algebra and designed for first semester students on bachelor level, is described. The learning tool has been built with three main modules: 1) Learning, containing audiovisual lessons based on sensorial stimulus, designed with the cognitive theory principles; 2) Reinforcement, where the understanding of the lessons is practiced using an approach based on games and student centered; 3) Evaluation, where the level of learned and practiced topics of the lessons is evaluated. The three modules development of the learning virtual environment are detailed, so as the auxiliary modules for the management of the information of professors and students, queries and reports.