Thu, 21 Aug 2008 17:25:35 BST CiteULike: parmentierf Smyth http://www.citeulike.org/user/parmentierf/author/Smyth CiteULike.org en-gb Copyright © 2004-2008 citeulike.org http://www.citeulike.org/user/parmentierf/article/838590 <i>(2006)</i><br /><br />We will demonstrate the topic model, a recent unsupervised learning technique that uses a statistical model to discover topics in a large collection of text documents. The first demonstration illustrates how the topic model automatically learns about the spectrum of research conducted by faculty members at UC Irvine and UC San Diego, how to topically characterize each researcher’s interests, and how to find researchers with similar interests – all in a completely unsupervised fashion. The second demonstration illustrates how medical researchers may use topic modeling to find new connections between genes and brain regions based on a large collection of articles on schizophrenia. Exploring Large Document Collections using Statistical Topic Models David Newman Arthur Asuncion Chaitanya Chemudugunta Vasanth Kumar Padhraic Smyth Mark Steyvers (2006) 2006-09-11T07:03:28-00:00 2006 automatic-learning data-processing statistics