Names are not perfect identifiers for persons. There may be two or even more persons with the same name. If only initials of some name parts are known, such a collision becomes more likely.
Most scientific papers are the result of collaborations of several persons. The composition of working groups, members of project teams, or groups of friends may change over time. But for persons with many publications it is very unlikely that the set of coauthors may be decomposed into disjoint groups of persons who never published together. If this nevertheless happens, our knowledge about the publications may be very incomplete, or there may be several persons with the same name whose publication lists where mixed.
Groups of coauthors connected by joint publications are indicated by the same color.Copyright © Sat May 16 22:54:16 2009 by Michael Ley (email@example.com)