You're getting confused between HTML/XML escaping and UTF-8/Unicode.
If the page is valid XML, life will be easier - you can just parse it as any other XML document, and then just get the relevant text nodes... all the XML escaping will be "unescaped" when you get the text.
If it's arbitrary - and possibly invalid - HTML then life is a bit harder. You may well want to normalize it into valid HTML first, then parse it and again ask for the text nodes.
If you can give us a more concrete example, it will be easier to advise you.
The HtmlDecode
method suggested in other answers may very well be all you need - but you should definitely try to understand what's going on first. For example, you may well want to only decode certain fragments of the HTML - if you decode the whole document, then you could end up with text which looks it contains like HTML tags, but actually just contained text in the original document.