In your travels around the internet, you have probably used many powerful search engines. Many of today's top search engines use the features such as those in interMedia text to provide the reliability of results that is expected from them. A case study that builds a search engine will be shown in a case study later in the module. If you were a computer book addict, and were looking for info, you could search for the phrase books. The problem is you will receive results on everything from cooking books to books on trains in no useful order.
By utilizing interMedia text, we can focus that search into a more useful result set. If you want to narrow down the search, you would specify a search like

books + potter

programming. This tells the search engine that you are searching about books, and want Potter books ranked highest, also you want to eliminate or lower the rank of results with programming books in the search.
Modifiers, such as + and , appear to be very useful in a search, but we must keep in mind that most people will not know that these special modifiers exist, and the average person will not want to spend time on reading through documentation or help.
Users will not usually spend the time to learn the ins and outs of proper usage. The average user will want to use natural language searches for 'themes' and 'gists'.
A natural language search is a search in the form of a question:

Where do I find out about books?

The theme of the previous question is 'books'. Oracle interMedia has integrated the ability to add to indexes themes of documents or concepts within a document, though it is up to you to plan the design around your intended user.