Information Addict

Having used a computer almost every day of my life for nearly fifteen years - as the only tool I know for organizing vast amounts of material from many disparate interests - I'm now wondering if we need better interfaces for the job. For me, the desktop metaphor is proving inadequate for keeping track of the complex relationships between large amounts of information.

See IneedaTerabyte?

Things are easier when the interface allows you to swap from overview to detail, view any stage between, create connections wherever they naturally occur, and - crucially - explore the result in a non-linear 'data landscape' that allows immediate, intuitive grasp of the relationships you're trying to discover.

This kind of consistent visual mapping is missing from the current desktop-based systems. However, so far, attempts to address this issue either:

require too much processing power for the average user;
are expert systems without universal application;
are not immediately usable via the web.

Work by Peter Young of the the Visualisation Research Group at the University of Durham is perhaps the most comprehensive to date:

http://vrg.dur.ac.uk/misc/PeterYoung/pages/work/documents/#IV-Survey

I did some experimenting with trial system using VirtualRealityModelingLanguage to map both quantitative and qualitative information from data on disability arts activity. The quantitative data (geographical location, links between organisations, type of organisation, level of funding, etc.) was used to draw primitive coloured shapes and lines representing connection between them. These were superimposed on a map of the region, with embedded links that brought up the qualitative accounts of each organisation. The 3D overview allowed users to grasp an organisations relationships, type, funding, geographical placement, etc. very quickly in one go, then drill down to get the details. The (now archived) site is:

http://arts-infoscape.org

-- DaveEveritt

I've found little CGI scripts helpful. I wrote a position paper describing one example for the OOPSLA'01 Software Archeology: Understanding Large Systems workshop. -- WardCunningham

http://c2.com/doc/SignatureSurvey/

The phrase when examining an unfamiliar program one needs to get a feel for the whole program at once (Ward, link above) seems to be the crucial part of any exercise in simplifying information - although how all this data fits together (Greg, below) may differ between individuals. The hard part is to chunk things in a simplified overview that provides an immediately-comprehensible context for each element, while offering the facility to go straight from each element in the overview to the details in as few steps as possible. -- DaveEveritt

I find the problem is not so much in organization, but in filtration. There is SO much information out there, but 90% of it is duplicated, or wrong, or phrased poorly... Maybe this is one of the selling points of a Wiki (yes, Dave, I'm coming round): irrelevant / overly verbose / incorrect information can be edited out. -- BenDaglish

Or, indeed, the filtration problem might be seen as the problem of how to convey understanding of the organization to the Information Addict. An example from my own experience is the JavaLanguage API. It is only my acquired familiarity with this API that enables me to find the particular method or class that I'm looking for. How can a computer instill in me a familiar understanding of 'how all this data fits together'? I think it must have its own, first. -- GregTurner

Is there an InformationAddictsAnonymous? group that we can join to help our problem?