This is an example of a malformed dictionary. It was created with the html entities where it shouldn't be. The problem is not in GoldenDict. In fact Babylon shows the same issue. Check the screenshots:
All that said, I'm trying to do a fix in GoldenDict for dictionaries with html entities:https://github.com/goldendict/goldendict/issues/53
Hopefully we could find a workaround. In the meantime you could use a tool such as UnpackBgl to extract a GLS file. This file is just a text file you can open in any text editor and do a find an replace of the html entities for the character you want. Then this corrected GLS file can be the input of the Babylon Glossary Builder program.
I'm not including any link here because I'm not sure of the legal status of these tools but you can find them for free by doing a search in Google. Good Luck!