Predicting Recovery Boiler Performance and Exploiting Data Visualization for the Critical Variables, 2004 International Chemical Recovery Conference

The epoch in which we are living has come to be known as “The Information Era” due to the gigantic volume of data currently accumulated by organizations. The current challenge is to transform these masses of data into information relevant to firms. Due to technological advances, on-line recording of variables during industrial operations is growing, so that is can now be done once a second, and the consequence of this is the generation of immense databases. To analyze such masses of data manually is impracticable, requiring the use of support tools and an appropriate methodology. Among other goals, we have data visualization and automatic techniques of variable selection.

The potential of data visualization is because of the human capacity to process visual information. Humans have the ability to rapidly map and identify images, being capable of detecting alterations in color, size, among other attributes, so that humans may be able to find out, by means of visualizing, tendencies and/or patterns among variables contained in databanks. (Marakas, 2003).