|ASIS&T 2006||START Conference Manager|
For example, consider three cited authors: A, B, C. If A is co-cited with B, and B is co-cited with C, but C is not co-cited with A, then there is also an associative (perhaps unknown) link between Authors A and C through Author B (A – B – C).
This method has been applied to journal keywords to determine associations previously unknown via the ARROWSMITH Project (Smalheiser & Swanson, 1998) but it also works well with other co-cited elements, such as authors. Moreover, this method of indirect contact can be extended beyond just a single intermediary. For example, given a chain of four authors, A – B – C – D, although each author is directly co-cited with whom they are directly linked, those not linked are not co-cited; and it may be of interest to know why and how—e.g., through which papers—those authors are related, e.g., how authors A and D are linked.
For example, if Pablo Picasso is connected to Joseph Stalin, it would be of great interest to an art historian or Russian historian to know which and how many intermediate authors and/or journals link them.
Although similar to social network analysis, this technique is unique in that it is applied through a co-citation framework, thus allowing for more general cited elements, such as cited journals, keywords, etc., and can be adapted for dynamic generation similar to that of other systems—notably, the AuthorMap System (Lin, White & Buzydlowski, 2001).
This paper explores this methodology by looking at a particular dataset, the Arts and Humanities Citation Index from the years 1988 – 1997, containing over 1 million records and a wealth of interesting chains (and answers the question as to whether Picasso is linked to Stalin).
|START Conference Manager (V2.52.6)|