576 JOURNAL OF LAW AND POLICY
themselves with which features appear relevant in a particular
case, as long as enough do. One such research project uses
cluster analysis to sort through a large corpus of emails and to at
least reduce the pool of potential authors from a great many to a
few with considerable accuracy.^99 Tim Grant’s work also makes
use of “a basket of features,” concerning itself more with the
number of features in common than with whether a particular
feature or kind of feature is likely to be predictive across
cases.^100
Whether I am right or wrong in predicting the subsequent
direction of the field, I firmly believe that far more collaboration
among scholars with different areas of expertise is absolutely
essential. And I would like nothing more than to see a
significant increase in collaboration notwithstanding any conflicts
experienced by the players and without a precipitating crisis.
(^99) Farkhund Iqbal et al., Mining Writeprints from Anonymous E-mails for
Forensic Investigation, 7 DIGITAL INVESTIGATION 56, 56 (2010).
(^100) Grant, supra note 96.