The aim of this study is to model students' academic performance based on their interaction with the online learning environment designed by researchers. The dataset includes 10 input attributes extracted from students' learning activity logs. And as an output variable (class) final grades obtained by students in Computer Hardware course was used. The predictive performance of three different classification algorithms were tested (Naive Bayes, Classification Tree, and CN2 rules) on dataset. Predictive performance of algorithms were compared in terms of Classification Accuracy (CA), and Area under the ROC Curve (AUC) metrics. All analysis were performed by using Orange data mining tool and models were evaluated by using ten-fold cross-validation. Results of analysis were presented as Confusion Matrix, Decision Tree, and IF-THEN rules. The experimental results indicate that the Naive Bayes algorithm outperforms other classification algorithms in terms of CA and AUC metrics. On the other hand models which are generated by Classification Tree and CN2 algorithm are easy to understand for non-expert data mining users.