There are no softwares to evaluate the spectral measurements instead of you. You need to identify species yourself. This is facilitated by the increasing number of spectrum prediction softwares, especially for NMR, IR. In case of the less frequently used Raman spectroscopy, as far as I am aware, you still need to rely on books and literature.
You can predict Raman spectra by DFT methods using one of the software available (a list can be found here http://en.wikipedia.org/wiki/List_of_quantum_chemistry_and_solid-state_physics_software) but as Imre said the identification relies on you. The best thing you can do, in my opinion, is optimize your molecules and run the harmonic analysis in order to get the IR and Raman frequencies. Once you did this you should compare the vibrational modes and assign them by comparing the experimental spectrum with the simulated one.
Firstly, If the symmetry of the molecule is known then a manual normal modes of vibration analysis can be done. Identifying irreducible representations and then ruling out the one's that do not correspond could be a good starting point. If access to high performance computing facility is available, then I believe one can perform a ground-state optimisation of the structure (run a single point calculation) and then also request for the frequencies to be shown. As suggested above by Luca, comparing this data with the experimentally obtained spectrum should help resolve most of the task.