Network embedding transforms a network into a continuous feature space where inherent properties of the network are preserved. Network augmentation, on the other hand, leverages this feature representation to obtain a more informative network by adding potentially plausible edges while removing noisy edges. Traditional network embedding methods are often inefficient in capturing - (i) the latent relationship when the network is sparse (the network sparsity problem), and (ii) the local and global neighborhood structure of vertices unique to the network (structure preserving problem). In this paper, we propose SENA, a structural embedding and network augmentation framework for social network analysis. Unlike existing social embedding methods which only generate vertex features, SENA generates features for both vertices and relations (edges) after solving the aforementioned two problems. We compare SENA with four baseline network embedding methods, namely DeepWalk, SE, SME and TransE. We demonstrate the efficacy of SENA through a task-based evaluation setting on different real-world networks. We achieve up to 13.67% higher accuracy for community detection and link prediction.