- Abstract: The tremendous success of deep generative models on generating continuous data like image and audio has been achieved; however, few deep graph generative models have been proposed to generate discrete data such as graphs. The recently proposed approaches are typically unconditioned generative models which have no control over modes of the graphs being generated. Differently, in this paper, we are interested in a new problem named Deep Graph Translation: given an input graph, the goal is to infer a target graph by learning their underlying translation mapping. Graph translation could be highly desirable in many applications such as disaster management and rare event forecasting, where the rare and abnormal graph patterns (e.g., traffic congestions and terrorism events) will be inferred prior to their occurrence even without historical data on the abnormal patterns for this specific graph (e.g., a road network or human contact network). To this end, we propose a novel Graph-Translation-Generative Adversarial Networks (GT-GAN) which translates one mode of the input graphs to its target mode. GT-GAN consists of a graph translator where we propose new graph convolution and deconvolution layers to learn the global and local translation mapping. A new conditional graph discriminator has also been proposed to classify target graphs by conditioning on input graphs. Extensive experiments on multiple synthetic and real-world datasets demonstrate the effectiveness and scalability of the proposed GT-GAN.