Big data analytics

Data mining - Extracting actionable insights from data

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 analytics, using sophisticated tools to present large volumes of data within statistical and visual models such as decision trees and three-dimensional graphics.

A typical CGI 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. CGI experts can conduct a data mining Proof-of-Value early in the BI life cycle to help identify which data (and combinations of data) will provide new insights and have predictive value.

A Proof-of-Value helps the organization construct data warehouses or data marts that will support future data mining efforts. Such initiatives can reduce future costs and project life-cycles by as much as 75 percent by dramatically reducing the time necessary for ongoing data collection and preparation.