Abstract: In order to overcome the problem of emotional deficiency in the traditional distance education,we proposed a learner's emotion prediction method based on fuzzy cognitive maps.The proposed method can be used to guide the teaching strategies adjustment effectively and help the learner become better engaged in the smart learning environment.Firstly,our method uses various affective factors as the inputs of fuzzy cognitive maps.Then,by using active hebbian learning algorithm,two emotion parameters are calculated,which stand for the important attributes of the emotion,i.e.,valence and arousal.Finally,the learner emotional state can be identified qualitatively and quantitatively by mapping the two parameters to a learner emotion space in two-dimensional Cartesian coordinate.The results show that performances of the proposed method are consistent with most students' real emotions,and the comparative accuracies with other Bayesian networks indicate the effectiveness of the fuzzy cognitive maps.