Hum, I hope that it's not for stealing/copying 
information without authorization, uh?You can use the library, or more simply
a filter that 
will process each html page (option -V) with some 
linguistical analysis, but I don't know any product 
that would do that