Information discovery in 3D environments: looking for compelling examples

12 Oct 2006 - 10:45am
7 years ago
2 replies
774 reads
Steve Toub


I work with some folks who are absolutely fascinated with 3D virtual
environments such as Second Life. Most of the applications we work on
are in the realm of traditional information retrieval: search engines,
library card catalogs, etc. While I have seen a few applications in
Second Life that provide information discovery (Google search, Amazon
book search etc.), in present form they seem to be a serious step
backwards from the traditional 2D search results screens that have
become common in the past 10 years: slower response time, harder to
navigate results, fewer results displayed, less information displayed
for each result, requiring to launch a traditional web browser for
delivery of the information object, etc.

Can anyone describe or point me to instances of compelling information
discovery applications that are well suited within 3D environments? Or
describe how 3D affordances improve upon 2D for information discovery?


Steve Toub


12 Oct 2006 - 1:35pm
Greg Petroff

Hi Steve,

Take a look at the enclosed links. I used to work for vizrt which has a very
impressive data visualization system installed at the NYSE. The original
interaction design and creative direction was a shared project between
NYSE's technology group SIAC under the direction of Dror Segal and Asymptote
Architecture. Vizrt provided the technology and integration services and has
since taken on the role of designing new features and integrating different
data streams. It is a fully navigable analog of the actual trading floor
with the mapping of transactions and the status of their IT infrastructure
in real coordinate space.

The NYSE is a good example as the real environment is a data rich space to
begin with and having a monitoring application that simulates that is very
helpful. 3D works well in this type of construct where geography is part of
the information space.

You can also take a look at the infovis comunity at for some
examples. But most of this work is academic in nature or focused on highly
complex data sets where adding another data dimension can help the expert


Gregory Petroff
Mobile # 646 387 2841

12 Oct 2006 - 2:44pm
Ben Hopkins

A long time ago, around the time when VRML was a cool new buzzword and
DirectX was still vaporware, I worked for a 3d modeling shop that received a
good deal of interest from chemical and electrical plants with the desire to
have their plants modeled. They wanted a singular 3d walk-through space to
help analyze and fix problems when things would break. They had a bunch of
scattered CAD diagrams that described individual parts of the plant but
without that navigable map it took alot of extra work to figure out just
where, say a gas leak was coming from and what were the major risk factors
in getting it isolated and then repaired.

More along the lines of searching non-graphical data, the business
intelligence/data warehousing conception of dimensionality is a good place
to start looking into how more complex data analysis could benefit from a 3d
data-space. BI basically defines a dimension as an orthogonal category of
information This type of dimensionality has no limits so its theoretically
just as easy to exceed 3 dimensions as it is 2. One common way of
representing dimensional data is with an OLAP
<>cube, which uses across and down
dimensions on a standard 2d grid to represent different dimensions and
heirarchical groupings of data within them, respectively. There's alot of
drilldown and shifting of perspective in navigating these cubes. This
necessitates reloading a new 2d grid every step of the way if you're stuck
in 2d space, but in 3d space you could see where you are in terms of
navigating these data dimensions in the cube and not be forever losing
context of what you had just drilled down from.


Syndicate content Get the feed