This volume is a selective post-proceedings of the 2015 IEEE/ACM International Conference on Advances in Social Network Analysis and Mining. The ten individual studies have been edited and considerably expanded beyond the original conference versions, averaging 25 pages in length (compared with fewer than 20 pages in most Springer conference volumes, and eight pages in the IEEE/ACM format of the original conference). The papers are uniformly of high quality and represent a good snapshot of the state of the art in the areas that they discuss. As expected in a conference proceedings, they reflect a range of different topics, but there are some overlapping themes around which we might organize them.

Three papers deal with the evolution of a social network through time. Social networks are often constructed based on user profiles constructed from their postings, and On-at et al. offer detailed experiments on the impact of taking the time of the postings into account in assessing the similarity between two users, compared with weighting all postings equally in constructing a profile. Harada et al.’s paper seeks to predict activity levels in a social network over time, a capability that might be useful (for example) in scheduling an online advertising campaign. Kaya et al. seek to predict future citations between scientific papers based on past ones, taking into account the times associated with the previous citations.

Two studies use methods of natural language processing to study network structure. Bharti et al.’s paper seeks to detect sarcasm, and Kang et al.’s paper uses n-gram analysis of microblog entries to distinguish those that are newsworthy (of broad potential interest) from others that are not.

Two papers deal with methods for influence maximization (identifying those nodes in the network with the greatest influence over the rest of the network). Gaye et al. study the problem of influence maximization in its own right, while Interdonato et al.’s paper applies influence maximization algorithms to detecting lurkers on a social network, with a view to encouraging them to contribute content of their own.

Two papers do not group naturally with the others. Fisher et al. evaluate a common assertion that social networks have higher degree correlation (one form of assortativity) than networks generated by other means, and find that this tendency is partially an artifact of the way in which network data is analyzed to reconstruct the network. Garcia-Martin et al.’s paper offers an energy analysis of the very fast decision tree algorithm, a computational breakthrough that is important in streaming processing of massive data (but is not at all specific to social networks).

The high quality of these carefully revised conference papers makes the volume of interest to researchers specializing in social networks who seek to stay abreast of recent developments.