Re: Quiz robot

[Rhandeev Singh <rhandeev@comp.nus.edu.sg>]

On Sat, 30 Jan 1999, the next to last of the red-hot fakirs wrote:
> >Why does the database have to be open to the students?
>
> Nobody will know if the people downloading the free, online
> databases are teachers making quizzes or students making ethical
> violations.  Shrug.

Short answer:

     It is not the database that needs to be kept secret;
     rather, it is the choice of questions.

Long, juicy answer with lots of implications:

With an open license, it is only a matter of time before nobody needs
to know any longer.

Imagine a database with 10 million multiple choice questions, a
hundred thousand open ended questions and fifty two and a half answers
to each open-ended question, and in 50 different languages (translated
automatically via something like e.g. the babelfish mechanism -- see
http://babelfish.digital.altavista.com).

If such a database is open to students, it can even be a teaching aid,
helping students to understand what the questions want out of them,
much like a ten-year series, only much better because it can be used
for web-based instruction, to build HOWTOs, to compare and contrast
alternative answers, etc...

Self-study becomes a viable option, and also a myriad of hitherto
unthought of modes of teaching and learning, many of which can at last
be automated interactively!  Even the database itself can receive
contributions from students (moderated by the database maintainers).

In fact, if a fellow student (yes, I'm a student too) is actually able
to remember all that information, I think he more or less deserves an
award for it :-)

Maybe, to raise awareness, we could one day receive advertising
revenue at the eventual database site and conduct online
multi-disciplinary quizzes where top performers get a "I've got
photographic memory" t-shirt mailed to them or something...

You see, it is not the database that needs to be kept secret; it is,
rather, the choice of questions.  And this can be kept secret via any
of a selection of well-known public key cryptographic techniques
designed into the applications that make use of the database.

I can see many legal issues, though, e.g. concerning the ownership and
maintainership of the database and its recognition, or lack thereof,
by relevant authorities... this in fact mirrors the analogous issue of
accountability of source code in the open source community.  Why do
the commercial elite like IBM and Netscape trust open source?

e.g. how do we prevent saboteurs from "polluting" the database?  If it
is completely open, then what are the chances that the good people in
the community will spot a mistake before it is used in some educator's
mid-term?

It is analogous to the issue of: how easy is it for a phracker to
create a backdoor in one of the standard Linux utilities and plant it
at SunSITE?  How many people will suffer from the security compromise
before it is detected?

(This is actually not so bad when we consider that each educator is
likely to check the questions and answers him/herself before
finalizing the term paper; in technical circles, on the other hand,
most system administrators take things for granted -- going through
millions of lines of source code is ridiculous!)

The IT industry seems to be handling this quite well, due to the
presence of bodies like the CERT (www.cert.org) and numerous other
organisations that contribute in many ways to the well-being of the
community as a whole.

My hope is that the education industry will demonstrate itself to have
a similar wealth of community-spirited people, as has been found in
the open source software community.

What we will need is for groups of educational institutions to combine
resources (e.g. over the Internet), and set up volunteer bodies that
inspect or otherwise moderate modifications to the database(s).

Educators who discover mistakes would be able to suggest corrections
to these bodies, and hopefully, the techies among them can try to
locate the origins of the errors (and discover phracks if any were
involved).

The law could have much to say about this.  What are the chances that
an educator could be sued in the States over a mistake he made setting
the mid-term paper?  Or marking the scripts?

What are the chances that he would make a mistake if he had used a
peer-reviewed database?  The chances could be much lower.  So maybe
trusting an open content database might be the right thing to do under
appropriate circumstances (e.g. given credible maintainership, a
history of dependability, etc).

---
Rhandeev Singh                          rhandeev@comp.nus.edu.sg
Linux User Group                        http://linux.comp.nus.edu.sg
School of Computing                     http://www.comp.nus.edu.sg
National University of Singapore