ANN: refdb-0.9.8-pre6 available

Hi all,

I've finished the work on the dupcheck branch and merged the result
back into the main line of refdb development. The result is a new
prerelease (built from svn revision 216) which you can grab right here
as usual:

http://refdb.sourceforge.net/pre/refdb-latest.tar.gz

There are two new features that make this prerelease worth a try:

1) Duplicate checks

refdbc received an additional command called checkref. The syntax is
very similar to the addref and updateref commands. In contrast to
these commands, checkref will add the reference data to temporary
tables. You can safely throw data at your database and see what
happens without running the risk of altering your existing data. In
addition to the output of addref which tells you whether or not the
data were added successfully, checkref will create a report of a
variety of duplicate checks. The format of the report can be selected
with the -A switch. "scrn" will return a text-only report, whereas
"xhtml" creates a more elaborate report viewable with your favourite
web browser. Moving around your mouse will pop up tooltips that try to
explain what to do with the match data. An example output can be
viewed here:

http://refdb.sourceforge.net/checkref.html

checkref tries to find duplicates of references by comparing the
titles, the location (journal name, volume, issue, start page for
journal, newspaper, and magazine articles), and the citation key.

checkref also tries to find similar author names, journal names, and
keywords. These matches are not primarily targeted at detecting
reference duplicates, but to give you a chance to keep your database
tidy and consistent. Typical cases are author names which are spelled
out in one reference but use abbreviated first and middle names in
another reference. If you can verify that the authors are identical,
you can expand the abbreviated name to the more useful spelled-out
name. Along the same lines, you can make sure that an abbreviated and
a full journal name are mapped to the same journal in the
database. Likewise, the keyword lists will help you to identify
similar existing keywords which may or may not be more appropriate
than the ones in the new reference.


2) Frequency data

RefDB can now help you to get a different view of your reference
databases. The getau, geted, getas, getjo, getjf, getj1, getj2, and
getkw commands of refdbc now support a "-s format" switch that lets
you request additional frequency information. "-s freq" will list the
absolute number of references that use the given authors, journals, or
keywords requested with the command. "-s relfreq" displays a relative
frequency indicator as an integer between 0 and 10 along with the
listed values. The formula used to calculate the relative frequency is
likely to require some more thought, and I'll be happy to receive
advice from people more apt in math than me.

The frequency data are even more fun with the (x)html output of the
getref command (this even works with the xhtml output of the new
checkref command). The relative frequency data are translated into
font size and color used to display the values, similar to the tag
clouds of del.icio.us and the like. Large author names or journal
names indicate that these authors or journals are represented above
average in your database. Likewise, a reference containing large and
red keywords is a "mainstream" article, dealing with the most
frequently listed issues in your database. A reference containing only
small and blue keywords would be considered less important in terms of
your average interests. Try this example output:

http://refdb.sourceforge.net/getref-frequency.html

To get the frequency display in the xhtml output, use the "-s KW"
switch of the getref command to add the keyword list to the output,
and the "-G /usr/local/share/refdb/css/refdb-frequency.css" switch to
select the new stylesheet that takes the frequency information into
account (you can make both settings permanent in your
~/.refdbcrc). The refdb.css stylesheet displays the references as in
the previous versions.

Both new features are likely to require some further polish. Please do
not hesitate to send your ideas and rants to the list.

regards,
Markus

-- 
Markus Hoenicka
[email protected]
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642