| 2 |
Thomson
Scientific’s journal coverage |
| |
Thomson
Scientific currently indexes approximately 10,000 journals in the
Sciences, Social Sciences, and Arts & Humanities. All journals
indexed by Thomson Scientific are peer-reviewed. As a group, the
Thomson Scientific-indexed set of titles represents an elite body
of internationally influential research publications. This set
does not represent a comprehensive catalog of all research
journals or all peer-reviewed research journals, but it is
generally sufficient to take into account all significant
journals.
Thomson Scientific indexes the segment of the journal
literature that exerts a disproportionate influence. The principle
involved in this coverage strategy is based on a well-known
concept in bibliometrics, Bradford's Law of Scattering. Bradford's
Law asserts that a relatively small group of journals will account
for the large majority of important and influential research in a
given field. |
|

return 
table of contents
|
| 3 |
Field
definitions |
| |
Individual
papers are divided into categories on the basis of the journal in
which the paper appeared. The fields used in Research Fronts
correspond to Thomson Scientific’s Current Contents®
(CC®) 106 categories. (See Appendix
1: List
of Categories)
Papers appearing in the multidisciplinary science journals
Science, Nature, and Proceedings of the National Academy of
Sciences of the USA (PNAS) were reassigned to specific
categories. |
|

return 
table of contents
|
| 4 |
Copyright
declaration and terms of licensing agreement |
| |
Thomson
Scientific retains international copyright to all data in Research
Fronts. The data are licensed to an individual or institution for
personal or institutional use only.
The data may not be
redistributed to any other party, resold in any form, or
incorporated into derivative databases. Publication—in
print or electronic form—of data residing in or compiled from
Research Fronts is expressly prohibited without the written
permission of Thomson Scientific.
If you have questions on
this licensing agreement, please contact the Research Services
Group, Thomson Scientific, 3501 Market Street, Philadelphia, PA
19104, USA. E-mail: nancy.bayers@thomson.com
(215-386-0100, x1776; fax: 215-387-1266); mary.vanallen@thomson.com
(215-386-0100, x1885; fax: 215-387-1266); or ann.kushmerick@thomson.com
(215-386-0100, x5351; fax: 215-387-1266).
|
|

return 
table of contents
|
| 5 |
Welcome
to the RSG Research Fronts database |
| |
The
Research Services Group of Thomson Scientific is pleased the
present this Research Fronts database and search interface. We
hope that you will find it useful for identifying and surveying
active areas of current scientific research, along with
statistical indicators of their intellectual and demographic
characteristics. If you have questions about this dataset please
contact: Nancy Bayers nancy.bayers@thomson.com, Mary Van Allen mary.vanallen@thomson.com,
or Ann Kushmerick ann.kushmerick@thomson.com. |
|

return 
table of contents
|
| 6 |
What
are Research Fronts? |
| |
|
6.1 |
Background |
| |
|
|
Research
areas in science, particularly those at the cutting-edge of their
fields, are characterized by patterns of intense communication
between scientists. This communication manifests itself in various
ways, both formally and informally; citations are one prominent
method of communication from one scientist’s work to another’s.
Patterns of citation reflect a fine-grained selection process of
how scientists build on each others’ work, and show the
relationship of these works to one another. Such patterns can be
used to create a picture of the state of a specific research area
in terms of the papers that constitute its core of seminal work,
and the associated set of citing papers that have, through their
referencing behavior, defined the core. |
|
return 
table of contents
|
| |
|
6.2 |
Definition |
| |
|
|
The
procedure to identify such patterns is called Research Front
analysis. It starts with identifying highly cited papers, and then
determining how often these works have been jointly cited. The
highly cited papers are jointly cited if the references/footnotes
of a paper include one highly cited item accompanied by another
highly cited item.
For the annual research front data, a single year (2005) of
citing papers is used. The source papers, those being cited, can
be from 1981-2005, spanning a 25 year period. |
|
return 
table of contents
|
| |
|
6.3 |
Process |
| |
|
|
Highly
cited papers are identified by a combination of integer and
fractional citation count thresholds. First, we select all source
papers that have been cited at least 6 times. Next, we use a
fractional citation counting method in which the references in
each citing paper are weighted by the number of references the
paper contains overall. For example, if a citing paper has 10
references, each reference receives a weight of 1/10. After
summing all fractional citations for a source paper, if the sum
exceeds a threshold of 1.0, the paper is selected as highly cited.
Fractional citation counts help eliminate the bias against low
referencing fields such as mathematics and engineering vis-ŕ-vis
high referencing fields such as biomedicine.
Next, we compute the number of times pairs of highly cited
papers have been cited together or co-cited. This defines the
frequency of co-citation of the two highly cited papers. To do
this, we set a threshold on the integer co-citation count to
eliminate very low values. Then, we convert the remaining pairs to
a normalized form using the following formula:
Normalized co-citation = Integer co-citation frequency of A and
B/(citation frequency A*citation frequency B)^.5.
This is the so-called cosine coefficient of similarity where we
divide the co-citation frequency by the square root of the product
of the citation frequencies of the two papers. A second threshold
is set on these normalized values. In this set of annual research
fronts a threshold of 2 or more was used for co-citation counts,
and the normalized threshold was set at 0.2.
Identifying research fronts involves manipulating the co-cited
pairs in order to group together those highly cited papers that
are strongly related. Starting with a co-cited pair that meets the
two thresholds, a clustering procedure then finds other pairs that
share common papers. The gathering or grouping process continues
until no other pairs of papers can be added to the set at the
selected thresholds. This process is commonly known as single-link
clustering. As an additional filter, if the resulting cluster
exceeds a size limit, the normalized co-citation threshold is
increased until the large cluster is broken up into smaller
constituent pieces none of which is greater than a maximum allowed
size. The resulting clusters vary in size from a minimum of two
papers to this maximum limit. In the current database, this size
limit was set at 50 papers. |
|
return 
table of contents
|
| |
|
6.4 |
Research
Front Metrics |
| |
|
|
Various
metrics for research fronts help determine the significance of the
areas and their stage of development. The number of core papers in
the front and the total citations received give an indication of
the size of the area. The number of citations per core paper
indicates the focus or intensity of effort. The average
publication year and distribution of core papers by year give an
indication of currency or "hotness"—that is, how
quickly research is changing and whether there are new
developments in the area. An analysis of frequently occurring
keywords or phrases in the titles of the papers, as given by the
front name, provides an indication of the subject content and
thematic focus of the area. If citing papers are provided in the
dataset, these papers provide further information on the current
state of the field and the cutting-edge of the research area.
Other indicators are author, country, institution, and journal
distributions. |
|
return 
table of contents
|
| |
|
6.5 |
Purpose |
| |
|
|
Research
front analysis is not designed to identify all research areas in
science or all papers associated with an area. However, it will
identify the most important research areas currently active in
science and the most important papers. It can assist in
identifying areas where current and important, highly cited work
is being carried out, areas where the scientific community is
focusing its attention, and the people and institutions most
active in each area. |
|

return 
table of contents
|
| 7 |
Opening
screen |
| |
|
7.1 |
Purpose: |
| |
|
|
The
opening screen shows all of the database’s research fronts. The
research fronts are listed from most cited to least cited. From
this screen, you can browse through the fronts and perform the
various functions described below. |
|
return 
table of contents
|
| |
|
7.2 |
Display
and Columns: |
| |
|
|
The
opening screen displays all of the research fronts in the
database: |
|

|
|
return 
table of contents
|
| |
|
|
There
are six columns of information (left column to right): |
| |
|
|
|
7.2.1 |
Front
ID: Numerical identifier for the research front |
| |
|
|
|
7.2.2 |
Research
Front Name: Description of the front including keywords used in
the fronts’ papers |
| |
|
|
|
7.2.3 |
Papers:
Number of papers in the front |
| |
|
|
|
7.2.4 |
Mean
year: Average year of publication of the papers in the front |
| |
|
|
|
7.2.5 |
Cites:
Total number of cites received by all papers in the front up to
the end of 2005 |
| |
|
|
|
7.2.6 |
Avg.:
Average number of cites received by a paper in the front
(cites/papers) |
|
return 
table of contents
|
|
|
7.3 |
Menus that appear at the top of the opening page: |
|
|
|
|
7.3.1 |
File |
|
|
|
|
7.3.2 |
Sort
by |
|
|
|
|
7.3.3 |
Run
Summary by |
|
|
|
|
7.3.4 |
Search
Research Fronts by |
|
|
|
|
7.3.5 |
View
papers for Front |
|
|
|
|
7.3.6 |
Help |
|
|
|
|
7.3.7 |
Select
window |
|

return 
table of contents
|
| 8. |
File
functions |
|
|
8.1 |
Purpose:
From the file menu, you can copy, print, save, and export research
fronts data. |
|

|
|
return 
table of contents
|
|
|
8.2 |
File
Options: |
|
|
|
|
8.2.1 |
Copy
to clipboard: copy all of the opening page information to your
clipboard |
|
|
|
|
8.2.2 |
Export
to excel: Export all of the opening page information to an excel
file |
|
|
|
|
8.2.3 |
Print
screen: Print current screen |
|
|
|
|
8.2.4 |
Print/save
full table: Print the entire table or save it to a file |
|

return 
table of contents
|
| 9. |
Sort
functions |
|
|
9.1 |
Purpose:
Enable you to sort the opening page information according to the
different columns. By default, the fronts are sorted by number of
cites. |
|

|
|
return 
table of contents
|
|
|
9.2 |
Sort
options: |
|
|
|
|
9.2.1 |
Avg.
Citations/Paper: Sort the fronts by average number of citations
per paper, in descending order. |
|
|
|
|
9.2.2 |
Cites:
Sort the fronts by number of |