From DNN 5.x, our website also began to use RadEditor as html provider for rich text editor, such as writing a post for our Ultimate DNN Blog Module – SunBlogNuke. Recently our clients noticed the issue that when typing characters in the Greek language it converts them to their html entities but not insert the actual unicode characters. For example, the sentence "Για να δουμε πως πα?ζει Ελληνικ?" will render "<p>&Gamma;&iota;&alpha; &nu;&alpha; &delta;&omicron;&upsilon;&mu;&epsilon; &pi;&omega;&sigmaf; &pi;&alpha;?&zeta;&epsilon;&iota; &Epsilon;&lambda;&lambda;&eta;&nu;&iota;&kappa;?</p>" in the HTML mode. With looking into the issue and request help in Telerik forum, we finally know that it is because the latest version of RadEditor included a new content filter named ConvertCharactersToEntities, which converts the special characters to entities, such as converts the Greek symbols to entities. In order to fix it, please open the configuration file and remove the ConvertCharactersToEntities string from the ContentFilters enum values, like that below:

<?xml version="1.0" encoding="utf-8"?>
<configuration>
  <property name="Skin">Default</property>
  <property name="AutoResizeHeight">True</property>
  <property name="ContentFilters">RemoveScripts,EncodeScripts,FixUlBoldItalic,FixEnclosingP,IECleanAnchors,MozEmStrong,
  ConvertFontToSpan,OptimizeSpans,ConvertToXhtml,IndentHTMLContent</property>
  ....
</configuration>

Hope it helps for somebody who encountered the same issue. :) Thanks again for Rumen and the Telerik team.

Refer to the original post mentioned the issue originally: RadEditor converting Greek characters to their html entities!