For more information on data warehousing
Knowledge Discovery techniques help organizations make sense of the large amount of data stored in their operational systems. CACI uses two key KD techniques - data mining and data visualization - to help organizations discover trends and patterns in their data and to help their employees make better, quicker decisions.
Most importantly, CACI helps translate the knowledge gained from knowledge discovery initiatives into business results. We define specific policies for key processes such as customer relationship management, credit risk management and database marketing.
We help our customers select and use the right tools to make critical business decisions. Our experts can design data mining solutions, conduct proof-of-value studies and technology surveys, as well as research new techniques and technologies.
Data mining is the extraction of actionable insights from data. Leading organizations use data mining techniques target profitable customers and likely prospects, reduce consumer fraud, identify loyal customers, improve customer service and increase customer retention. Data mining often involves predictive modeling, using sophisticated tools to present large volumes of data within statistical and visual models such as decision trees and 3-D graphics.
A typical CACI data mining project includes an iterative approach toward the following high-level activities
- Understand the business problem
- Prepare the plan
- Select the techniques, tools and platform
- Evaluate data sources and acquire data
- Prepare data for analysis
- Mine the data
- Interpret results
- Implement models
- Assess model performance and business impacts
Data mining is best executed as part of an integrated business intelligence (BI) effort. An organization's BI strategy will identify questions appropriate for data mining, data marts, data warehouses or analytical applications. CACI experts can conduct a data mining Proof of Value (POV) early in the BI lifecycle to help identify which data (and combinations of data) will provide new insights and have predictive value.
A POV helps the organization construct data warehouses or data marts that will support future data mining efforts. Such initiatives can reduce future costs and project lifecycles by as much as 75% by dramatically reducing the time necessary for ongoing data collection and preparation.
CACI helps identify patterns in data by applying the research efforts of data architecture experts. We take large volumes of business and customer records and apply sophisticated statistical and visualization techniques (such as scatter plot matrices and parallel coordinate plots in conjunction with brushing, cutting and high dimension rotation) to present a new perspective.
Data visualization involves more than drawing a picture. It also can provide a multi-dimensional image that can be rotated and twisted to uncover relationships among groups of data, versus focusing on individual occurrences.
By presenting data in such dynamic ways data visualization techniques often enable analysts to discover new trends or patterns in less than half the time of traditional statistical techniques. More important, because analysts can "see" their data and follow how the patterns emerge they are likely to have greater confidence in the meaning and reliability of the results.