Information for competitive advantage at Axel Springer publishing house
| Date | 01 February 1995 |
| Pages | 143-144 |
| Published date | 01 February 1995 |
| DOI | https://doi.org/10.1108/eb045353 |
| Author | Simon Hamilton |
| Subject Matter | Information & knowledge management,Library & information science |
Brief
Communication
Information for competitive
advantage
at Axel
Springer
publishing house
Information for competitive
advantage
at Axel
Springer
publishing house
Simon Hamilton
Information Dimensions
Ltd,
16th
Floor
Centrepoint,
103 New Oxford
Street,
London
W1A,
UK
What do Claudia Schiffer and Helmut Kohl have in common? Their biographical
details,
as far as
they
are
available,
are recorded in the
electronic
text
archives of the
Axel Springer publishing
house,
Germany's
largest
newspaper publisher. Whether
it is Ms Schiffer's hobbies or Kohl's food preferences, a wealth of information
about
them
and other famous people can be called up on the
computer.
This infor-
mation has been compiled from a wide
variety of publications
by the archive
librari-
ans. The problem is, how do you make an archive which contains more than 30
million press clippings accessible
to
users?
This biographical database exem-
plifies the changes Axel Springer has
undertaken since the mid-eighties.
Where information used to be ar-
chived and
forgotten,
now information
and documentation have been recog-
nised as a key factor in the success of
the organisation. Under the supervi-
sion
of Beate
Pauluth-Cassel,
about 80
staff members supply approximately
2000 editors at the publishing house
with extensive factual information.
The biographical database is the
most concentrated and
frequently used
information available in the archive.
More than 30 million clippings from
daily, weekly and monthly newspa-
pers are collected here and put into
endless rows of coloured lacquered
steel cupboards. Political events and
land,
sea, air and natural catastrophes
are to be found here
alongside
the
mar-
riages of European aristocracy.
Previously, even simple enquiries
had
to be handled by specially trained
staff,
who also had to be well ac-
quainted with the design
of
the
archive
structure.
The ever-increasing number
of enquiries, however, made easier ac-
cess essential.
'An important aim of
the
installa-
tion of
the
electronic text archive was
to enable editors to carry out simple
enquiries from their
desk,'
Joachim
Neubert, database organiser for the
text archive, explained. 'By reducing
their workload, information special-
ists were able to devote more time to
complicated individual and
back-
ground
queries.'
In
1987
the
development of
the
bio-
graphical database
and a database
with
information on films and television
broadcasts was started: since 1988
work has been progressing
on
the fac-
tual database which makes the rest of
the
files
accessible.
Full-text retrieval plays only a
small part when working with docu-
ments.
The quantity of archived infor-
mation meant that pure text research
would lead to vast numbers of hits,
making evaluation of the selected ma-
terial too expensive. For this reason it
was
far more
important
to
implement a
keyword search index which would
quickly locate the most relevant docu-
ments.
A completely new classification
was developed for the whole archive.
Keywords
and
synonyms were organ-
ised in a tree-like structure which can
be further developed when necessary.
In this way
a
relatively
small,
control-
led vocabulary
can be used to
describe
the contents of every document and
file, and it is possible to create a rigid
frame into which all text files fit.
'The classification is a powerful
means of cataloguing existing docu-
ments,' Joachim Neubert explains.
'However, on its own this is not suffi-
cient. We cannot take for
granted
that
all users know all of
the
14 000 key-
word chains by heart.'
For this reason
a second
method of
searching was created —
a
thesaurus
with newspaper-specific vocabulary.
This thesaurus consists of general and
subheadings which help
the user to
de-
rive
the
keyword.
It also has cross
links
and alternative terms which help the
user during
the
query.
'While searching for the right pro-
gram we were faced with a problem,'
Joachim Neubert remembers. 'On the
one
hand
we
wanted to
record the data
in a clearly structured
pattern;
on the
other
hand
we needed the ability for a
free text query. The program also had
to be able to represent our whole the-
saurus.'
Due to the special data pattern —
partly unambiguous fields, partly free
text —
both
typical relational data-
bases and retrieval systems were
con-
sidered.
However, databases lacked
the capability for enquiries; retrieval
programs were not able to cope with
structured
data.
Classification of the search terms
not only provides access to the docu-
ments but is also essential for
an
opti-
cal record of
the
documents. The ar-
chive is growing by 3500 articles a
day, which is
causing an
enormous or-
ganisational expenditure in
man
hours
and takes up a lot of space — some-
The Electronic Library, Vol. 13, No. 2, April 1995 143
To continue reading
Request your trial