The doctoral thesis behind the 2008 method. Its title names the problem the rest of this record addresses: exact inference that is not bounded by treewidth, the structural limit that stalled the field.
Mark Chavira's doctoral dissertation at UCLA, completed in 2007. It is the source the NASA Ames paper cites for the compilation approach it used on ADAPT, and the conference papers from 2005 to 2007 in this record are the published work of the same years.
Treewidth is the structural property that put exact reasoning out of reach for real systems. The dissertation's title is the thesis of the company in four words.
A public copy has not been located online, so this entry is citation only. The 2008 journal paper is the published account of the method.