In this paper, we propose a method of obtaining the sense of touch by using the EEG. Multimodal device based on human sensory systems have obtained a lot of attention in the fields of human interface. Among the human sensory systems, especially, the sense of touch is important form of social interaction and it can have powerful emotional consequences. Therefore, it is important to improve tactile sensation device technique and evaluate the device objectively. By analyzing the EEG, we attempt to make the evaluation method clear how users feel when he/she touches objects. We measure the EEG while subjects touch objects by using portable electroencephalograph. The proposed method consists of feature extraction and EEG pattern classification by using fast fourier transform(FFT) and artificial neural network(ANN). Moreover, we apply particle swarm optimization(PSO) for deciding the weight and bias of the ANN fastly.