Galaxy interactions and mergers are a major driver of cosmic star formation, even if their contribution to the SF budget in the Universe compared to quiescent star formation in isolated disk galaxies is still unclear and debated.
Gas compression in galaxy interactions and mergers results from two mechanisms: a global one (radial gas inflows towards the nucleus) and a local one (compression and fragmentation of the ISM in turbulent flows with a Mach number well above one). The latter mode was really probed only recently, although observations of triggered ISM turbulence in interactions have existed for more than a decade. Numerical simulations are hence barely starting to model merger-induced starbursts in a realistic way, and cosmological models probably do not account for these events in a realistic way, which will be a crucial aspect to understand the global star formation history of galaxies.
Nevertheless, recent accurate observations of ISM properties and star formation in mergers can be accounted by theoretical models. These include a "double law" of star formation for mergers versus isolated disks, unveiling a "starburst mode" and a "quiescent mode", as well as an excess of dense molecular gas in the ISM of starbursting mergers. This probably shows that our theoretical understanding of the driving processes has improved and that modern simulations can model these in a realistic way. Major issues nevertheless remain, such as the role of stellar feedback processes, the triggering of gaseous outflows, the influence of gas accretion by central supermassive black holes in mergers, the mergers of such black holes into bigger and bigger black holes but with gravitational recoil effects, and the energetic feedback from these black holes.