see http://mindprod.com/jgloss/unicode.html

At the bottom is a link to a searchable text file giving a description
of every Unicode character.

write a little program to extract what you need and create either a
HashMap or byte[] whichever turns out to be more compact.

To make the whole thing more compact, you might encode some of the
repeating patterns in the descriptions as bit encodings.

If you don't need high speed, you could put it into an SQL database.

?Is there a method that returns a string containing the name of a
Unicode character?  I could not find one. Thanks, Alan?

I don't know of Java, but i know easy solutions is emacs and Python...
that might be helpful...

http://xahlee.org/emacs/emacs_n_unicode.html
http://xahlee.org/perl-python/unicodedata_module.html

 Xah
 xah@xahlee.org
 http://xahlee.org/