I mentioned in Optimal Representation of Text Documents that a tag cloud can be used to illustrate high frequency terms in a document but doesn't show any real structure of that document. One way to improve this is to position the tags in a cloud so that tags which are used together in the document appear close together in the cloud. Tag clouds usually show their terms in alphabetical order or are sometime sorted in increasing order of frequency.
Tag clouds are ordered the wrong way: Tags denote concepts. As such, they have meaningful relations to each other. Tag clouds are ordered alphabetically or by size - it would be much more effective, if tags that belong together could also be presented together. Some of these relations can be deduced automatically, by observing how tags are used: Some tags might always appear together, others sometimes and others never. If tags co-occur frequently or have many common 'neighbors', you can be sure the concepts denoted will be related in some manner.
Here is an example of his that gives the idea. It's interactive so be sure to play with it to get the full effect.
Colour is used here to show the 'freshness' of a tag, whether it was used a lot recently or not. When you click on any tag it enlarges it as well as other tags which are highly related. This gives a pretty good way to explore a fairly dense collection of words and see how they relate to one another.
It might be fruitful to apply these ideas to the important words within a document rather than a set of tags used across a collection of documents. Perhaps colour could be used to indicate relative position within the document so that key terms which appear more frequently near the beginning would be more green. Another idea would be to use a topic-based colouring scheme similar to what I did with Topic Flowers.