SnowDeal is aggregating recent articles about text mining. A good statement of the problem is provided by BioNLP: The literature of the field of biology is the largest of all the sciences. The volume of biology literature each year, measured in bytes, is about fifty times the size of the entire human genome, junk and all. But locked in this literature is an enormous amount of information that can tell us much about the structure and function of genes, proteins, cells and organisms -- how they work as well as how they can fail.
Very interesting stuff.