Building on these answers, VASP does not calculate IR or Raman activity itself, but you can use a pair of scripts to do it using the eigenvectors from a DFPT phonon calculation (or a finite-difference one, using IBRION = 5/6):
IR intensities: http://homepage.univie.ac.at/david.karhanek/downloads.html#Entry02
Calculating IR intensities requires the eigenvectors and the Born effective-charge tensors, which can be obtained from VASP using the LEPSILON or LCALCEPS tags - usually a fairly straightforward calculation.
Calculating Raman intensities is more difficult, as it requires two further calculations per mode (i.e. if you have N atoms in your structure, you need to do 3 * N * 2 further calculations). You can of course use a symmetry analysis to determine which modes *should* be Raman active, and only calculate the intensities for those modes.
Regarding symmetry assignment, Phonopy (http://phonopy.sourceforge.net/) can give you the irreducible representations of the mode eigenvectors for some point groups, but not all.
By using a combination of these programs, you should be able to obtain all the information in the paper Mohamed posted, although as you might have gathered CASTEP has a better infrastructure for phonon calculations than VASP does.
In one of our papers, we simulated the IR and Raman spectra (including estimating the linewidths) of Cu2ZnSnS4 using VASP - this might be interesting to you if it's what you need to do: http://scitation.aip.org/content/aip/journal/aplmater/3/4/10.1063/1.4917044.
MedeA Phonon utilizes the Phonon/PhononA software of Prof. K. Parlinski to automatically calculate IR and Raman spectra of solids using VASP (https://info.materialsdesign.com/Datasheets/Datasheet-Phonon.html).
I've found many time the mention of this site [https://homepage.univie.ac.at/david.karhanek/downloads.html#Entry02] to obtain a way for IR analysis, but it seems this web site is not longer supported. So, I just wanted to ask if anyone has the script that is supposed to be found in this site and can share it. I am doing some spectral analysis and want to take a look to this procedure.
For Raman spectra calculation you need additional software to work with VASP. For example, you can use PHONON software made by Prof. K. Parlinski. http://www.computingformaterials.com