In response to my post on taxonomies and folksonomies, Matt Hodgson offered a solution of topic maps as a way to deal with the unstructured chaos that comes along with folksonomies.
I have been looking at a similar approach of leveraging topic maps in my search for an enterprise search engine. Products like Recommind are able to create a relationship among words and phrases in a document and equate them as synonyms. So when you search for a "P & S" the search engine will return a "purchase and sale agreement", "sale agreement", and "purchase agreement" as well as a"P&S."
The next step is to get them to recognize the folksonomy tags and create relationships among them as well as the words in the document. Effectively, the folksonomy tag would be just another piece of metadata for the document.
