EEG,as a direct response to brain activity,can objectively reflect a person's emotional state.However,the non-smoothness and complexity of EEG signals make it difficult to collect a large number of labelled EEG samples,thus limiting the effectiveness and generalization performance of EEG emotion recognition methods.Here,a Semi-Supervised Low-Rank Representation (SSLRR) approach for EEG emotion recognition is proposed to address the above issues.First,an objective function in regression form is designed using the estimated labels of a small number of labelled EEG samples to effectively estimate the labels of unlabelled samples.Second,an ε-drag-and-drop technique is used to ensure label-to-label separability,and in addition,low-rank constraints are imposed on the slack labels to improve their intra-class tightness and similarity.Then,a class neighborhood graph is incorporated into the proposed approach to capture the local neighborhood information of all EEG sample data.Comparative experiments are conducted on two public datasets of SEED-Ⅳ and SEED-Ⅴ,and the results show that the proposed approach performs well in EEG emotion recognition.