The Author-ity Tools Homepage

Author-ity: tools for identifying Medline articles written by a particular author

displays clusters of papers written by the same individual,
is very fast (clusters are precomputed by an automated prediction algorithm),
is comprehensive (contains all names on all papers in MEDLINE/PubMed),
is static (only includes articles published up to a certain date)
- Author-ity 2008: includes PubMed and MEDLINE up to ~September 2008.
- Author-ity 2006: includes articles in the baseline 2006 version of MEDLINE.
permits further annotation and drill-down using PubMed and Anne O'Tate tools.

Alternatively,

Author-ity 2008 can be accessed directly from PubMed when using the Firefox Greasemonkey script for Author-ity
The complete Author-ity 2006 dataset is available, upon request, for non-profit academic research (under these conditions).

References:

Torvik VI, Smalheiser NR. Author name disambiguation in MEDLINE. ACM Transactions on Knowledge Discovery from Data 2009; 3(3):11.
Smalheiser NR, Torvik VI. Author name disambiguation. In Annual Review of Information Science and Technology (B. Cronin, Ed.), Vol. 43 (2009), pp. 287-313. A preprint version of this chapter is available here, and the final print version is available from the publisher Information Today, Inc.

displays a pairwise ranking of articles by similarity to a given index paper,

is slower (rankings are computed in real-time),

is up-to-date (includes the most recent articles in PubMed as well as old ones),

gives the estimated probability of author match for each pair of papers,

gives an accurate relative ranking,

however, the probability estimates do not include all correction factors employed in the clustering dataset.

We welcome and appreciate any questions, comments or suggetions via email to Vetle Torvik.

Last modified: Sun Feb 22 2009