You mean you want all the text from any tags in any structure under the <Education> tag? That sounds rather strange, but would be best handled by a push parser. Something like XML::Parser where you specify a subroutine to be called for start tags that will check the tag name and increment a global flag if it's Education, another to be called for text that'll push the text into a global array if the flag is set and last one for the closing tags that'll check whether the tag name is Education and decrement the flag.

Most efficient and in this case actually fairly simple.

On the other hand ... what are you REALY trying to do? We know this one step, but maybe if we see the bigger picture (know the task in which this is just a step) we can give a better advice.