Hi:
Generally speaking, SOURCE tables are the INPUT to a DI Studio process or transformation. A lot of times, SOURCE tables might be legacy files, such as SAS files, mainframe files, database tables or other types of data sources.
TARGET (not "taget") tables are the OUTPUT from a DI Studio process or transformation. You usually want/need the TARGET tables to be defined in the metadata so other users (people using EG, WRS and the SAS Add-in for Microsoft Office) can access the target tables. Or, it's possible that the DI Studio process would build an OLAP cube or load into a datamart or data warehouse.
It is up to you to design the structure of the libraries for your TARGET (or OUTPUT) tables. Some installations design libraries by department (Accounting, Sales, Purchasing, etc) -- other installations design libraries the fit other functions -- Analysis libraries, Reporting libraries. Really, that planning process needs to be part of the job that your Platform Administrator or DATA Administrator would do.
When you say "user defined libraries", it implies to me that you mean your users will issue LIBNAME statements. Generally, once you have a target table in a library that's defined to the Metadata, you need to decide what kind of security is going to be used for that library. Depending on what you set up, the users never issue LIBNAMES to the data -- they access the data through EG or WRS or AMO -- which shows the user only the tables that they are able to read/work with.
If, by user defined libraries, you mean that the Data Administrator has to set up the libraries -- then, that is correct. You would use the New Library Wizard to set up a user defined library to hold your target tables.
Reading the Platform Administration documentation, the DI Studio user's guide or taking some of the BI Platform Data Integration classes might help you with these tasks.
cynthia