The revolutionary discoveries of the last few years have opened a new era of astronomy. With the detection of gravitational-waves, we now have the opportunity of investigating new phenomena, such as mergers of black-holes. Furthermore, multi-messenger observations now allow us to combine information from different channels, providing insight into the physics involved. With this rapid evolution and growth of the field, many challenges need to be faced. In this thesis we propose three data analysis strategies to efficiently study the coalescences of compact binaries. First we propose an algorithm to reduce the computational cost of Bayesian inference on gravitational-wave signals. Second we prove that machine-learning signal classification could enhance the significance of gravitational-wave candidates in unmodelled searches for transients. Finally we develop a tool, saprEMo, to predict the number of electromagnetic events, which according to a specific emission model, should be present in a particular survey.