SEARCHER'S VOICE
The
Critic
by Barbara Quint
Editor, Searcher
Magazine
How would you like to make a living being nasty? (Why
are you all looking at me that way? SHADDUP!!) If I
may be allowed to continue (harumph), what this world
needs is more critics in place, people who can evaluate
the quality of information on the fly, people who can
continually monitor established sources, people who
can accurately rate the reliability of information
before its use puts their clients' interests and welfare
in danger. Clearly, in most situations, effective critiquing
of information will require specialized knowledge and
expertise, the kind possessed by specific types of
knowledge workers. But only information professionals
have the generalized knowledge to detect the patterns
of development or non-development that determine the
authenticity of data sources. Only information professionals
can assess the full information needs of clients and
correlate those needs with the specific content of
identifiable sources.
The other day I was working on an Infotoday.com NewsBreak
about a new service called WorldData. The international
economic and financial data in WorldData comes from
an alliance of three leading business news services:
the Economist Intelligence Unit (EIU), EcoWin, and
Alacra. These are all respectable suppliers of economic
data and, like all good hunter-gatherers in this subspecies
of the online industry, are always on the look-out
for new or better data sources. In fact, these services
proudly display their latest captures by announcing
them in "New Data Added" or "Latest On..." sections.
However, scan through such announcements and you quickly
realize how erratic coverage must be for users trying
to do comparative studies. For example, the press release
that initiated my research into the WorldData service
boasted that it provided country profiles for 150 countries
with real-time data. However, the fastest data deliverer
in the alliance — EcoWin — states on its
Web site that it only covers 80 countries. And that's
just a broad brush example. Wait till you try getting
into the nitty gritty of which service carries housing
statistics for what country and at what level of detail
and for which years of coverage.
Text services have equivalent problems. The other
day I did a search in Factiva's Publications Library
and retrieved a set of headlines and annotations for
a client. Ten days later, the client sent me back the
search and asked me to retrieve a subset of full-text
articles. (Yes, Whipper-Snapper Searchers of the World,
there are still clients who hire intermediary searchers
to do such things.) All the articles popped out as
quick as you please, except one. For some unexplained
reason, an article from the Miami Herald had
disappeared. A trip to the Herald's own Web
site and to Dialog's archive also failed to turn it
up. But did that deter the relentless searcher? Do
you really need to ask? And I did find the article.
Where? Sitting there with a bow in its hair, cute and
perky, in a Google News cache. Another $2.95 picked
from the pockets of commercial sources! Upon perusing
the article, the only reason I could discover for its
disappearance from official archives might be the tag
saying it came from a Washington Post news feed.
But Google had snatched it from the World Wide Web's
waste basket and made me and my client happy.
Speaking of The Washington Post, from what
I heard, it was the only one of the major newspapers
that yawned its way through the Tasini crisis.
The Post has always kept its digital archive
free of articles not written by their own staff or
under work-for-hire contracts that granted permission
for electronic re-publication. That solves one publisher's
problem, but what about searchers looking for the missing
articles? And, even worse, what about searchers — end
users and professionals — who perform a search
on the Post's archives and just assume the search
was comprehensive? Some embarrassing scenarios come
to mind. Can't you just see Senate or presidential
aides standing in front of their masters, heads hung
in shame, as clippings are waved under their noses
and tirades delivered with all the roar of the final "Trump"-et
("You're fired!")?
Sometimes you can't even rely on such basic human
motivators as greed to secure data flows. The other
day, while on Amazon ... again (sigh), I decided to
check out an author for whom I had not yet received
an announcement from the Amazon Alert service that
provides e-mail notices about new or forthcoming books
correlating to a user-supplied list of authors or titles.
This particular author has produced books every 2 years
for about a decade and I had gotten kind of worried
not hearing from her. To my surprise, the author had
released a book dated that very month. Why on Earth
had Amazon failed to send me a message?! Usually Amazon
Alert notices go out well before the book is published
to promote pre-ordering. A series of messages to Amazon
revealed that it had ceased the Alert service, with
no notification to customers or chance for customers
to retrieve their lists. At this point, I continue
to receive "people like you" recommendations, but the
marketing geniuses at Amazon appear not to recognize "Me,
Myself, and I" as qualifying as "people like you." Apparently,
Amazon tossed the Alert lists without integrating the
information into its other recommendation databases.
Why a company that so aggressively pursues business
opportunities as to build databases that cross-correlate
purchases by one buyer with those of others in the
hope of finding items that might attract a purchase
would ever cavalierly refuse to accept specific requests
from buyers defies logic. Even the logic of greed.
But I'll tell you one thing. I am now looking for a
book service that will supply such notices and, if
I find one, I will definitely keep a file copy of my
favorite authors, just in case.
Insuring data quality requires constant monitoring.
It's hard enough to train end users in general rules
of protection, like checking home pages as well as
pages embedded deep inside a site or learning to separate
sponsored sites from un-sponsored ones in Web engine
searches. It would be impossible to expect end users — even
assuming they had the skills — to evaluate the
quality of individual sources and data streams. Staying
on top of data quality issues is hard enough for information
professionals, but at least info pros have the connections
needed to do the job. For one starters, they have each
other. They share concerns and discoveries on listservs
with fellow info pros. They also know key vendor personnel
and how to approach them. Even if they don't know the
best person with whom to speak when they start an investigation,
they can very quickly find the right name by checking
with other colleagues. Actually, sometimes dealing
with vendors can work out better if you use a name
supplied by colleagues, as in "I was on the BusLib
list the other day trying to find out if anyone else
had had the same trouble with your files as my client
had. Several people told me they had. I promised to
get back to them after I talked with you. Now when
did you say you'd get back to me on this?"
See! I told you being nasty could get you places.
...bq
Barbara Quint's e-mail
address is bquint@mindspring.com.
|
