ALG: Visualization of Global Soybean Production







	Login

Overview

Quality Assurance Methods for Processing Microarray Imagery

Using Visualization for Microarray Data Analysis

Visualization of Soybean Production

Contacts: Loretta Auvil, lauvil@ncsa.uiuc.edu
Collaborator: Steve Sonka, UIUC; Donna Fisher, UIUC; Randall Westgren, UIUC
Funding Source: Bureau of Justice Statistics

Visualization of global soybean production

Problem Definition

Visualization can dramatically improve and speedup data comprehension by providing the means to interactively analyze data. Through manipulation of the visualization tool, users can also more readily identify patterns, trends and outliers in the data. In this project, the Automated Learning Group collaborated with the National Soybean Research Laboratory (NSRL) at the University of Illinois to develop a tool for visualizing projections of global soybean consumption for the years 2001 to 2025.

The Protein Consumption Dynamics (PCD) system uses a Powersim model, developed at the NSRL, that evaluates population and income growth data to predict regional protein needs and malnutrition. The model tracks the estimated annual consumption of six agricultural commodities---beef, pork, poultry, fish, oil/fats, and vegetable protein---serving as protein sources for humans in eight regions that together comprise the globe. The system is designed so that alternative scenarios of the future can be examined using population and income projections of the World Bank and the UN's Food and Agricultural Organization.

Approach

The visualization component of the project was developed by the ALG using the tool In3D Java from Visible Decisions Inc. Because of the nature of the problem and the type of data, a geographic metaphor formed the basis for this visualization. A world map was used in which eight regions were depicted by color value as per the specifications of the NSRL. A back "wall" perpendicular to the map provides additional information regarding the population and income for these eight regions. 3D bars were also used to indicate protein consumption per region as well as to provide a means for comparing consumption between two regions. Specific features of the visualization tool include:

Brushing

A number of brushing operations augment and provide more detailed information about the represented variables in this application. Brushing the mouse cursor over individual countries provides protein consumption, and malnutrition statistics for that country. Brushing the 3-d bar in the regions indicates regional protein consumption and will give a numerical breakdown by protein type for that region.

Comparison

Regional comparisons can be made by clicking through region names and years in the bar chart at the far-left of the visualization space. Clicking on a region on the map and a date below the map will change the bar chart to the right of the one just described. This provides easy comparison of regions and dates.

Change of View

This visualization application also allows the user to choose from a number of different views or scenarios. For example, when looking at the range of years, located at the bottom of the visualization space, users can click on the Title (e.g., Historical Data) and change it to one of three different scenarios of future consumption generated by the simulation model. Clicking on the blue spheres in the application will change the view of the back wall, the map space and the bar chart on the left.

Results

Through formal experiments with actual managers in the soybean sector, the effects of the PCD system on individual and group decision-making processes are being systematically evaluated.