How big the level of comfort and level of satisfaction of students in participating in learning needs to be known by the organizers of learning to improve the quality of learning and increase student interest in learning. One of them is by providing a facility in the form of an Academic Guidance System that students can use to express their complaints and opinions regarding the learning activities carried out. In this study, the author will classify with a text mining algorithm on any complaints or opinions given by students. The method used in this research is Lexixon Based Sentiment Analysis using two tools, namely Flair NLP Framework Application and KNIME Application. These two tools can classify every complaint or opinion into two categories, namely positive and negative. Positive can mean that there are no problems related to the lectures being carried out while negative means the opposite. The results of the application of the text mining algorithm on the KNIME Workflow that were applied in this study succeeded in mapping positive and negative sentiments from student learning, with an accuracy of 70%, still less than the desired standard of 85%.