Information Technology Reference
In-Depth Information
Fig. 7.2
A
Venn
diagram
showing
potential
links
between
bibliographically
unconnected
literatures (Figure 1 reprinted from Swanson and Smalheiser (
1997
))
In parallel, Swanson also published his work in the literature of library and
information science, notably (Swanson
1986a
,
b
; Swanson
1987
,
1988
,
1990
). The
Venn diagram in Fig.
7.2
, adopted from (Swanson and Smalheiser
1997
), shows sets
of articles, or bodies of literature, the target literature A and the source literature C.
Set A and set C have no articles in common, but they are linked through intermediate
literatures B1, B2, B3, and B4. Undiscovered links between A and C may be
found in through the intermediate literatures B's. There may exist an intermediate
literature B such that a particular transitive relation can be established based on
A
!
Bi and Bi
!
C.
Figure
7.3
shows a schematic diagram of title-word pathways from a source
literature on the right (C terms), through intermediate title-words (B terms), to title
words of promising target literatures on the left (A terms) (Swanson and Smalheiser
1997
). A ranking algorithm ranks discovered A-terms. The more B-pathways an A
term has, the higher it ranks. Term A3,
magnesium
, is the highest ranked title word.
It has a total of 7 pathways from B-terms. In this way, a pathway from
migraine
to
magnesium
appears to be most promising.
Swanson called this algorithm procedure I. Swanson also developed what he
called Procedural II, in which titles from literatures A and C are downloaded first
in order to find words and phrases in common from the two literatures. Common
words and phrases are selected to form the so-called B-list. An output display
is then produced to help the human user compare A-titles and C-titles against
B-terms.
Figure
7.4
shows B-terms selected by Swanson's Procedure II for magnesium
and migraine, and for fish-oil and Raynaud's disease. The two numbers in front of
Search WWH ::
Custom Search