Mining, Analysis and Interpretation
Discovery-Driven Systems Reveal Important Hidden Data
Apr 10, 2009
The following tasks need to be completed to make full use of data stored in the data warehouse, query generators, as well as data mining techniques and other business tools, to develop customer profiles:
- Create prediction and classification models
- Analyze links
- Segment databases
- Detect deviations
Creating Models
This task makes use of data warehouse contents to generate a model that predicts desired behaviour automatically. Discovery-driven models generate accurate, comprehensive models that are comprehensive because of their sets of if-then rules. For example, in an investment environment, the discovery-driven model can predict the performance of a particular stock to assess its suitability for an investment portfolio.
Analyzing Links
The goal of the links analysis is to establish relevant connections between database records. An example is the analysis of items that are usually purchased together, such as a washer and dryer. Such analysis can lead to more effective pricing and selling strategies.
Segmenting Databases
When segmenting databases, collections of records with common characteristics or behaviors are identified. One example is the analysis of sales for a certain time period, such as President’s Day or Thanksgiving weekend, to detect patterns in customer purchase behavior. This is an ideal task for a data warehouse.
Detecting Deviations
Detecting deviations is the opposite of data segmentation. The goal is to identify records that vary from the norm, or lie outside of any particular cluster with similar characteristics. The discovery from the cluster is then explained as normal or as a hint of a previously unknown behavior or attribute.
Modeling Techniques
There are several modeling techniques that aid data mining efforts. These techniques include:
- Creation of predictive models
- Performing supervised induction
- Association discovery
- Sequence discovery
Creating Predictive Models
The creation of a predictive model is facilitated through numerous statistical techniques and various forms of visualizations that ease the user’s recognition of patterns.
Supervised Induction
With supervised induction, classification models are created from a set of records, which is referred to as the’ training set.’ This method makes it possible to infer conclusions from one set of descriptors of the training set to the general. In this way, a rule might be produced that states a customer who is male, lives in a certain zip code area, earns $25,000 to $30,000, is between 40 and 45 years of age and listens more to the radio than watching TV, might be a possible buyer for a new camcorder. The advantage of this technique is that the patterns are based on local phenomena, whereas statistical measures check for conditions that are valid for an entire population.
Association Discovery
Association discovery enables the prediction of the occurrence of some items in a set of records if other items are also present. For example, in the healthcare sector, it is possible to identify the relationship among different medical procedures by analyzing claim forms submitted to an insurance company. With this information the prediction could be made -- within a certain margin of error -- that for a specific treatment protocol, the same five medications are usually required.
Sequence Discovery
Sequence discovery aids the data miner by providing information on a customer’s behavior over time. If a certain person buys a VCR this week, he or she usually buys videotapes on the next purchasing occasion. The detection of such a pattern is especially important to catalog companies, because it helps them target their potential customer base more effectively with specialized advertising catalogs.
These analytical tools enable stored customer data to be used effectively in the enhancement of customer relations and to increase customer profitability.
 |
