Creating a Virtual World for the Visualization of Data

Posted May 12th, 2010 at 3:42 pm.

Marissa Mocenigo and Priscy Pais
Mentor: Dr. Dianna Xu

The human brain is highly skilled at visual pattern recognition; for example, a pie chart is often much easier to comprehend than a list of numbers. Our research builds on this fact to develop a new method of visualizing data through the creation of 3D virtual organisms called protobytes, which will reside in a virtual ecosystem. The base code of the protobytes would be able to take in data, irrespective of the source or type, and create aesthetically pleasing organisms to represent the data. The changes in the lifestyle and health of the protobytes will represent alterations that have been made in the data and/or possibly in the ecosystem.

Working in conjunction with Professor Ira Greenberg of Miami University, our research aims to create a 3D creature, modeled after Greenberg’s own 2D protobytes, using Java 3D and Greenberg’s Processing software. The organisms will be composed of simple shapes and patterns in order to create a more primitive, basic appearance, much like a jellyfish. Protobyte anatomy also includes internal systems which control morphology, motility and behavior. They are bred utilizing evolutionary algorithms in conjunction with environmental influence. Individual protobytes can have their own autonomous behavior, competing in a dynamic system. Ultimately, a protobyte will be capable of movement; moreover, within the virtual ecosystem, it will be able to explore, become sick, heal, age or even create offspring through mating. This visual system would be useful in representing data in a variety of disciplines. For example, in the stock market, organisms of different colors or shapes could represent various stocks. The increasing price of a stock could be shown by an organism growing in size. A merger between two stocks can be depicted by two different protobytes combining together to give rise to a protobyte of a different structure.

The creation of the ecosystem is among our long term goals, as the project is relatively new and largely undeveloped. As such, our short term goals include creating the structure and overall appearance of various protobyte families and determining how the data will be interpreted by the system. In writing our code, we hope to create an extensive, extendable base class that allows for future developments in the creatures and the ecosystem without having to rewrite a significant portion of the code.

Comments are closed.