Bibliomics
Towards a New Model for
Textual Information Representation
Mindtel, LLC
October 2001
1.0 Introduction
In the Fall of 1996, the Defense Advanced Research Projects Agency (DARPA), in a BAA entitled Pathogen Countermeasures, solicited proposals from researchers working in everything from biochemistry and immunology to chemically resistant war fighter uniforms to information and communications technologies. It was a monumentally ambitious interdisciplinary program with nothing but the highest level of relevance and global urgency. Our group at that time had been asked to submit because of our innovative work in the areas of human-machine interface development and information representation. We submitted and were funded for the development of one of the core technologies essential to the processes of gathering diverse forms of data, integrating them into a unified matrix comprised of the critical data streams, and finally presenting them to an individual situated within a high performance information/communications query and response system we refer to as a ‘perceptualization environment.’ Since then, we have only continued to push for the development of novel interfaces.
The conceptual and methodological basis for our work is the principle of constructing information in ways which capitalize on the powerful features of the human nervous system for perceiving and rendering to consciousness different kinds of information. We believe that while information technology has exploded in terms of speed, volume of data processed and contained, and capacity for manipulation, what has been left out of this progress is the human. If technologies are not developed which allow the human user to more fully benefit from the abilities of the hardware and software, outcomes will suffer, especially in those areas with the greatest risk for human costs. Bibliomics is a critical area of human-machine interaction research with which we propose to close the gap between the human and the horsepower.
2.0 Bibliomics and Information Representation
Bibliomics is a research are which aims at the problem of the growing volume of electronically stored data and its inaccessibility to those in need of it. At the time of the Pathogen Countermeasures program, we were confronting the challenge of infected and threatened war fighters whose infection needed to be identified and, ideally, neutralized in a very short amount of time. Experts tasked with responding to such a crisis thus needed a highly integrated information gathering and representational technology. This technology required the capacity to
- collect
data from chemical and biological sensors in the region of the detonation
- run
this data through filters which would then go out into the highly diverse
global array of relevant data bases to rapidly assemble information
leading to the identification of the germs and/or chemicals released, and
- representation
of this vast amount of information in novel perceptualizations to allow
the expert to quickly make decisions facilitating eventual defense of
ground personnel and others affected.
The vision of Bibliomics is to develop technology that will search, find, and represent information. The representation of information in highly concentrated perceptual, in this case visual, forms is our current objective.
2.01 Visualization of Information
While our eventual plan is to include multiple sensory systems in the novel representation of information, for now we are focusing on the visual system. In short, the idea of Bibliomics is relatively straightforward. Staying with the example of biological and chemical warfare[1], lets say a rocket has exploded and has thrown up a cloud of chemicals and germs among a large group of war fighters. Sensors are rapidly detecting these elements and sending their raw data to a central command space. The operator therein then sends the results of this data in the form of a query into a thousand databases all over the world. As the special search engines begin to stream their data back to central command the user has no time to read through tens of thousands of pages of complex information to decide how to preempt the action of the lethal agents. Our proposal is to change all of that data into non-textual forms which will allow a more rich and powerful decision impending interaction with information.
In our work to date, we have developed a set of rich geometric shapes which we call torroids. These are both simple and complex objects whose attributes of shape, size, color, texture, dynamicity, morphability, and so on, we hypothesize to be rich sites for the encoding of information. In particular, we aim to develop a set of rules for this new “geometric alphabet” which facilitate rapid signification and assimilation. Instead of seeing lines of text, the user will see basic shapes emerge, complexify, aggregate, and form “sentences” in real time as millions of bits of information are collected from these distributed data stores. We are thinking essentially of a new kind of literacy. The user would be trained to see relevant information in these highly concentrated meaning rich dynamic objects. By the time the objects emerge for the user, they represent stored data which has undergone a radical reduction in the “noise” with which they are saturated in native form. Through this process, a decision maker may respond to a crisis more quickly and powerfully than ever before because a greater amount of information has been brought to bear on the emergency thus allowing a more potent response.
3.0 Conclusion
Mindtel seeks the opportunity to move forward in the development of the bibliomic approach to the representation of information. Currently, we are working with some prototypes of the idea described. We are convinced of the urgency and relevance of these ideas for both obvious and not as yet obvious reasons. Below are examples of this technology. As they stand they are arbitrary. The idea is to develop such objects in collaboration with specific users so as to intentionally and richly load them with significance as perceived by local users.
