1) Files .dz are compressed with gzip standard. you need to rename them as .gz in order to extract the files.
2) Extract the .dict file from the archive in the same directory where the .ifo file is.
3) Use stardict editor for windows:
http://code.google.com/p/stardictpro...1.rar&can=2&q= to convert the .dict file in plain text (in linux usually it's in the package stardict-tools), the utility needs the .ifo and the .dict files in same directory.
4) the resultant text file is probably structured like this:
word
<tab character>definitions
each word in a single row.
With a text editor ( or sed, awk in linux) substitue the tabs with
2 spaces, and you may erase most of xml tags, they're not used by converter.exe (with sed I usually change characters coded in html format with unicode characters using a small script).
5) your dictionary is ready for conversion.
the text file needs utf-8 coding to be used by converter, so better using text editors supporting that coding, for small text files I use openoffice writer or gedit.