In recent years, the deepening of the role of emotions in education has increased, since it has been shown that emotions influence what we learn and retain, in different areas such as learning a second language. Likewise, the analysis of emotional data in learning environments has been implemented through various methods, such as the collection of vital signs, where brain-computer interfaces provide a means to capture emotional metrics from brain activity. Furthermore, public data sets in this domain are scarce. Consequently, this work presents the initial approach to generate an Emotional Dataset, for emotional analysis based on the collection of emotions, through an Emotiv Insight as a Brain Computer Interface. Also, through the Babbel application for learning second languages. As a result of this study, it will be possible to obtain the general and projected structure for the resulting data set, based on the generated approach.