XML data is self-describing. That means it contains both data and information about the data. In records of traditional database systems, before you store data, it requires to define relational schemata, file description tables, external data definitions etc. Where as in XML, these things are not required. Because the data itself contains all these information. This is very important for seamless integration of data, as far as business applications are concerned.
XML can be used to describe and identify information accurately and unambiguously, in a way that computers can be programmed to ‘understand’ your information (well, at least manipulate as if they could understand it).