One of the data mining problems receiving enormous attention in the database community is classification. Although Artificial Neural Networks (ANNs) have been successfully applied in a wide range of machine learning applications, they are often regarded as “black box”, that means predictions cannot be explained. To enhance the explanation of neural network, a novel algorithm is to extract symbolic rules from neural network has been proposed. With the proposed approach, concise symbolic rules with high accuracy can be extracted from the trained neural network. Extracted rules are comparable with other methods in terms of number of rules. The network is first trained to attain the desired accuracy rate. Redundant connection of the network are then removed by a network pruning rule. The effectiveness of the proposed approach is clearly demonstrated by the experimental results on a set of benchmark data mining classification problems.