[Sigia-l] Findability

[Boniface Lau]

> From: sigia-l-admin at asis.org [mailto:sigia-l-admin at asis.org]On
> Behalf Of George Olsen
>  
> On 1/27/03 5:03 PM, "Boniface Lau" <boniface_lau at compuserve.com>
> wrote:
> > IMO, hand-crafted categorization is just too expensive and is
> > heading towards extinction!
> 
> And why are information retrieval and categorization mutually
> exclusive?

Not sure why you take "hand-crafted categorization is just too 
expensive" to mean "mutually exclusive".


> 
> I'm working with a financial services company on an extranet for
> institutional investment clients. These folks' interest is in
> efficiency.  They'd _much_ rather click on links that take them to
> the latest research reports than have to enter search queries.

In addition to getting the latest research, what if your clients want
to have reports with certain attributes? Are you going to create an
"attribute maze" for your clients to go through?

Efficiency is indeed important. But going through a maze, albeit with
the nice-sounding name "categories", is hardly an efficient way. It
reminds me of lab mice discovering their way through a maze. What a
torture!

In data access terminology, navigating through a maze is "sequential
access"; retrieving through a search mechanism is "random access". 
Care to guess which access method is more efficient?


> Especially when the categories were based on understanding how they
> thought about the information themselves.

Mind you, the way people think about a piece of information changes,
depending on their experience and what they want to accomplish at 
the moment.


> 
> If I forced the clients to use a Google interface, they'd scream
> because it's shifting the cognitive overhead to them.

As if navigating through a category maze does not require cognitive
effort? 

Navigating through a category maze created by someone else is a
second-guessing game. It is like living in someone else's house. While
you know clothing items should be in the closet but knowing where the
socks are requires some hit and miss discovery.

In particular, when navigating through a category maze, people have to
constantly guess what the category maker meant by the various category
names. Furthermore, when they guess wrong and therefore make the wrong
turn, they may not realize that until they have made many more
subsequent turns along that wrong path. Thus, they have to backtrack
and go down another path. But they may not even know at which point
they were making the wrong turn. Thus they have to start all over
again. Isn't that frustrating?

Now, contrast that with the effort required to come up with search
terms for a Google-like interface, isn't the Google-style a breeze?

Effort aside, we all know that students who haven't done their
homework prefer multiple-choice questions over essays during an
exam. Should we assume web users, in your case institutional
investment clients, are really that ignorant about their interest area
to prefer multiple-choice over coming up with search terms?

In spite of your claim that your categories were based on an
understanding of how users thought about the information, deep down do
you really believe that anyone understands how others thought about a
piece of information? Haven't you forgot about the recent debates here
on exactly what information is? When there is so much disagreement on
even the very basic, what give people the confident that they
understand how others thought about that disputed basic, i.e.
"information"? Therefore, isn't forcing users to navigate through our
category maze a bit egotistical?


Boniface