As explained in What are the basic ClearCase concepts every developer should know?, the main difference you will find between ClearCase and most of the more recent VCS is:
ClearCase reasons file-by-file, and not on the repository level.
So when ClearCase makes a baseline, it actually takes all the latest versions of the files of a given component, and apply a label, for each file.
SVN will simply make an atomic operation, making a new revision of the repository with a new tag (which is actually a cheap copy in a tag "directory", like SVN branches: see "What do you use the svn tags directory for anyways?")
Note also that "baseline" in ClearCase refers to the UCM methodology, which is a complement to ClearCase, and which has no correspondance in UCM.
A baseline in ClearCase is for a "component", ie a specific subset of all the files of a VOB.
An SVN repo is just a massive centralized place where you can version any number of files you want. You can consider a specific directory of that SVN repo as a component (and "tag" just that), but that is entirely at your discretion: you won't "declare" a component in SVN first, before "baselining" it.