Since the spread of Covid-19, online learning is considered the main solution for universities so that the lecture process continues to run well, but in online learning there are many obstaclesfaced by students and lecturers, as for the obstacle, namely the location of the house is not covered by the internet network, student behavior is difficult to understand. monitored and absorption of material is very lacking and the obstacles faced by campus are the difficulty of knowing how big the level of student satisfaction in online learning is, on this problem, the author wants to evaluate the online learning system by first classifying the level of student satisfaction with online learning using the C4.5 algorithm. , based on the results of data processing that is deemed less convincing, data processing is also carried out using another classification method, namely the Naive Bayes method. Based on the results of the comparison of the two algorithms, the accuracy of the algorithmNaïve Bayes is better than C4.5 by a difference of 11.77%. That the accuracy rate of the algorithmC4.5 is 58.82% with the validity test using cross validation obtained the average value for the value of Class Precision Dissatisfied 0%, Class Precision Satisfied 66.67%, Class Precision Unsatisfied 0%, Precision Class Very Dissatisfied 0%. And the value of Class Recall Dissatisfied 0%, Class Recall Satisfied 90.91%, Class Recall Unsatisfied 0%, Class Recall Very Dissatisfied 0%. While the level of accuracy using the AlgorithmNaïve Bayes with the results of 70.59% with the validity test using cross validation obtained the average value for the value of Class Precision Dissatisfied 0%, Class Precision Satisfied 76.92%, Class Precision Unsatisfied 33.33%, Class Precision Very Dissatisfied 100% and Class values Recall Dissatisfied 0%, Class Recall Satisfied 90.91%, Class Recall Less Satisfied 25%, Class Recall Very Dissatisfied 100%.