Visualizing Online Information
By Paula Hane
Tim Bray has had a 20-year career in the software industry. He is recognized
as an expert in Web architecture, information retrieval, and software optimization.After
managing the New Oxford English Dictionary project at Canada's University ofWaterloo,
he co-founded Open Text Corp. and introduced what would become one ofthe first
commercial Web search engines. In 1996, he joined the World Wide Web Consortium's
XML Working Group, serving as co-editor of the eXtensibleMarkup Language (XML)
1.0 specification. In 1999, he founded Antarctica Systems and is now the company's
CTO. In this interview, Bray talks about Antarctica's progress in advancing
the role of visualization software for improving users'access to online information.
Q: The last time we spoke was about a year ago when
Antarctica was testing a prototype graphical interface for the DIRLINE database
for NLM. (See the NewsBreak at http://www.infotoday.com/newsbreaks/nb020513-1.htm.)
What has happened with that?
A: The goal we have all been shooting for is to do
something for TOXLINE, NLM's toxicology database, but we haven't gotten started
on that yet. The prototype hasn't been developed any further, but there are
people using it. NLM is not a real fast-moving organization. But the feedback
on the prototype has been good, and I'm confident we will move forward on the
Q: Have you done other work with NLM or other government
A: We are in discussions with several organizations
within the Department of Defense, but there's nothing we can talk about yet.
The DoD has been doing quite a lot of aggressive work in the KM space recently.
It's interesting, though, that there's not much actually happening in KM right
now in the private sector.
Q: At one point, Antarctica had several interesting demos available
at map.net, including a PubMed demo site. I couldn't find them now.
A: Go to http://maps.map.net and
you'll be able to see them. Last July,Antarctica hired a new CEO. We decided
that our technology was a little too horizontal, a bit too general. We were
having trouble when we showed prospective customers our very generic demo.
We actually lost a few pharmaceutical contracts because the demo didn't speak
to them. So now when we talk with a potential buyer, we try hard not to lead
with product, but instead try to figure out the problem the company is trying
to solve and what data resources it has. When we actually show them a demo,
we invoke a demo that speaks their language. It's a matter of sales strategy.
We prefer to lead with questions and then follow up with a demo.
Q: I understand that your CEO, Barry Yates, is a seasoned sales professional.
It makes sense for a technology person to bring in someone else to lead.
A: Yes, I've done this before. With my first company, Open Text, I
was CEO for the first 18 months and then hired a replacement. And we went on
to do a big IPO and have fun. I would hope that pattern repeats itself. Unlike
a lot of technology people, I have the highest respect for people with expertise
in the sales and marketing arena. I understand that it is at least as demanding
as building technology.
Q: Let's talk about some of the specific markets you are now targeting.
One of them is close to my heart, the library market. Antarctica recently signed
a distribution agreement with Neptune8.com, a software consulting firm that
specializes in libraries. Neptune8.com is now bringing your Visual Net product
to the library, library automation, and research markets in North America and
A: There were just too many libraries for us to reach. We had been
making some headway, but it seemed inefficient for a small company in the Pacific
Northwest to attempt to market to the whole library community. The head of
Neptune8.com felt there was real potential for Visual Net in the market, and
we are happy to work with them. He forecasts two deals for this quarter.
We will also look to work with other distributors. The natural choice would
be the ILS vendors. We know those people and they know us. I think some are
rightly watching to see how some of our engagements turn out. We are also working
with one of the library technology organizations on some projects that should
produce a lot of credibility in the library space.
Q: Your first installation ofVisual Net at a library was at Belmont
Abbey College (http://belmont.antarcti.ca).
How has this worked?
A: We've had an excellent experience there. Don Beagle, the library
director, has published a couple of papers with good, quality-controlled research
about usage by undergraduates.
Q: So the company could present real data to potential libraries about
a payoff or return on investment (ROI) for implementing this technology. The
real problem libraries seem to have right now is losing patrons to Google.
It seems that a tool like Antarctica offers could help libraries fight back.
A: You should come to work for us. That is almost exactly, word for
word, the value proposition we've been pushing to libraries. But we have had
a bit of a problem in that many librarians think they are rather well-served
by their current toolsets and are disinclined at this point to make investments
in bold new directions. But their customers are less well-served, which we
think is the issue.
Q: Libraries are also struggling with their budgets, given the difficult
economic times we are in. These days, I'm seeing absolutely must-have technologies
that are struggling to find buyers. Everything is a tough sell right now.
A: We've tried to be brutally disciplined in our key direct-sales
efforts, only going after deals where we could make a really convincing business
case in dollars and centshere's how much you pay, and here's what you
get. We've been doing quite a bit of work in what we call the business intelligence
For example, we're doing a deployment now with a large valve manufacturerreally
unglamorous stuff, faucets and such. The company has a major task in managing
its inventory worldwide and its database is very large. It's hard to think
of a large manufacturer having points of commonality with a library catalog,
but it does. We've allowed them to drill down through the company's inventory,
by where inventory is sitting, how long, the volume, etc. So instead of asking
a programmer to write and run a report against the database, they can navigate
through the data by point and click. This gives them a very easy-to-calculate
ROI. They can reduce their labor costs by so much per week and reduce their
inventory by so much. And we cut them a deal for the software that pays for
itself. During tough times, we have to be looking at that kind of scenario.
Q: What other business-intelligence applications would work, besides
catalogs? What about for enterprise document management or knowledge management?
A: I'm totally convinced that is one of the key growth areas for us
going down the road. The trouble is I see a lot of buying paralysis out there
in regard to KM tools. As we become increasingly an information-centric economy,
the importance of KM will become more and more strategic. The companies that
have what we call a good enterprise memory are the ones that will win. But
at the moment, people tend to be focused on tactics to the exclusion of strategy.
You've got to meet payroll, right? So at the moment, we're not marketing aggressively
in the KM space, since it's in transition and somewhat troubled.
Q: What about partnering with companies that are in the enterprise
search space? Verity and Autonomy hold a large chunk of that market, I believe.
Could you see Antarctica working with some of those companies?
A: We've done work with some companies in this area. The way our technology
works, it would be very easy to snap on. We see Visual Net very rarely deployed
in stand-alone solutions, so our success will largely be in these kinds of
Q: The one area in which I've noticed some activity is taxonomy and
A: We've done some deployments with Semio, and the match is very good.
To the extent that this becomes a hot growth area, we'd be delighted to make
our splash there. Every one of them has a nice XML export format that makes
it easy to work with. The problem is, as a discipline closely related to KM,
there's not a lot of action right now. There are also a huge number of auto-categorization
vendors, and it's not clear which ones are going to win, so we've been a bit
reluctant to invest too many resources. The partnerships we're working on right
now are actually with ERP systems. [They're] not nearly as exciting as KM systems,
but these are the applications that track inventory systems. And a better interface
can clearly translate into ROI.
Q: Antarctica calls Visual Net a "browse engine" and said that it
is not trying to replace search engines but can work as an enhancement to them.
Have you targeted any of the popular general Web search engines for partnerships?
A: A search engine is great if you know what you are looking for,
just like a library catalog will find a book for you. However, neither of them
do a good job for browsing contentthe equivalent of browsing through
the library stacks. That's where we feel we do a great job. There are good
synergies with search engines. We have had conversations with Google and some
other well-known engines, but we have nothing to announce yet.
Q: With the search engines trying to outdo each other, adding
enhancements and bells and whistles, do you think adding visualization functionality
could be a next step for them?
A: Yes, absolutely. The basic search results list creates an illusion
that the result is one-dimensionalit all lines up in a row from most
useful to least. But in fact, in a sufficiently large database, any result
should be multidimensional. Our tool provides a useful categorization of the
results and makes it visually apparent.
There's a big gap. Everyone has visual interfaces on their desktops, but
as soon as you get into a shared information space, you deal with lists of
text. I think the search business is just getting started.
Q: What about your competitors? There's WebMap, TheBrain.com, Inxight.
How do you differentiate from these others?
A: There's another new one, called Groxxis. We differentiate Antarctica
in using visualization techniques from cartography. We use map-making tools.
In addition, our technology can be patched onto just about anything in Web
services. We don't really see these companies, though. Our chief competition
is the CFO who doesn't want to spend any money.
Q: Let's talk about an announcement you plan to make in the next few
weeks. Which? Online, the Internet arm of Consumers' Association, a U.K.-based
consumer research organization, has licensed Antarctica's Visual Net software
for four of its popular online guides: The Good Food Guide, The Good Bed & Breakfast
Guide, the Which? Hotel Guide, and the Which? Guide to Country Pubs.
This type of consumer application seems to me to be a logical fit. Users
can interact with maps to locate hotels, pubs, and restaurants, etc. Will you
try to market more to consumer sites?
A: We weren't smart enough to dream this up ourselves. They came and
found us. And yes, other sites would absolutely be potential targets for us.
Our plan is this: First, we have to do a really great job with Which?make
that application a slam-dunk, make them happy, and meet their business needs.
These online guides are subscription sites, but we'll be able to show it to
We are doing some research into the power players in the online guide space.
Then, we will do some marketing. But the best marketing will be to have a happy
Nobody's getting rich quickly these days, but if you have a quality product
and integrity in how you sell it, there's lots of room for good growth out
For more information on Antarctica Systems and Visual Net, visit http://antarctica.net.
Demos are available at http://maps.map.net.
Paula J. Hane is Information Today, Inc.'s news bureau chief
and editor of NewsBreaks. Her e-mail address is email@example.com.