Well, if they wanted you to be able to get at the text within the CDATA node, they should have given you:

<AccomodationName><![CDATA[HOTEL AC MALAGA PALACIO]]></AccomodationName>

If they had, the text content of the node would be:

“HOTEL AC MALAGA PALACIO”

Instead, the double-escaped it so that the CDATA tag is part of the text:

“<![CDATA[HOTEL AC MALAGA PALACIO]]>”

Since you are getting an opaque text node which just happens to contain some CDATA syntax, you will need to re-parse the text node if you want to extract the text from within the CDATA brackets.  To do this, you would need to extract the string, then pass it back through another XML parser instance which is capable of handling fragments.  You could also theoretically just extract the inner part using string manipulation routines or RegExp.