Traditional machine learning and data mining algorithms mainly assume that the training and test data must be in the same feature space and follow the same distribution. However, in real applications, the data distributions change frequently, so those two hypotheses are hence difficult to hold. In such cases, most traditional algorithms are no longer applicable, because they usually require re-collecting and re-labeling large amounts of data, which is very expensive and time consuming. As a new framework of learning, transfer learning could effectively solve this problem by transferring the knowledge learned from one or more source domains to a target domain. This paper focuses on one of the important branches in this field, namely inductive transfer learning. Therefore, a weighted algorithm of inductive transfer learning based on maximum entropy model is proposed. It transfers the parameters of model learned from the source domain to the target domain, and meanwhile adjusts the weights of instances in the target domain to obtain the model with higher accuracy. And thus it could speed up learning process and achieve domain adaptation. The experimental results show the effectiveness of this algorithm.