A Vision on Social Content Retrieval

Many years of research and exciting achievements in the multimedia information retrieval (MIR) field have made the field mature enough to en ter a new development phase – the phase in which the technology enabling automated indexing a d relevance ranking of multimedia content is steered towards adoption in practical re trieval scenarios and applications. Although the retrieval performance of these machine-level se mantic inference solutions will always involve trade-offs in precision vs. recall, many of such solutions are already providing value for users, and many significant hurdles in developm ent of MIR technology have been passed [8].