Data mining is the practice of searching through data sets to find patterns. The term can be somewhat of a misnomer as it draws comparisons to underground mining – which involves the extraction of valuable material from a larger deposit of unwanted material. 

In data mining, however, all the material collected can be va How is data mining completed? 

 Most commonly, the approach follows the Cross-Industry Standard Process for Data Mining (“CRISP-DM”), an open-standard process, which involves 6 steps: 

 

• Business Understanding – What are the organization’s objectives that can be supported by data? What are the goals of using data? 

• Data Understanding – What data assets does the organization have? What needs to be collected? What preliminary relationships can be noticed? 

• Data Preparation – What segments of the data should be selected for this process? How should the data be prepared? To what extent does it need to be cleaned?  

• Modeling – How should the data be modeled to achieve the organization’s objectives and goals? What design considerations should be made?  

• Evaluation - How do the different models compare? Did the models achieve the organization’s objectives? Should we move forward with deploying them? 

• Deployment – Build a report to share insights from the data set. Continue to monitor and maintain the models.
  

This process will ideally produce models that find patterns, relationships, outliers and correlations within the selected data set. 

 What are the benefits? 

The effectiveness of data mining depends on the goals and objectives being set at an organizational level. The rise of big data offers incredible potential for organizations that can use it strategically. 

Data mining can help organizations make better decisions. By analyzing data sets, patterns will emerge that can help you make better predictions about future events, recognize outliers and improve forecasting efforts. Data mining can bring relationships between variables into focus. For example, analysis can help identify a segment of a business’ market that they may not have been targeting. 

It can also help you identify weaknesses in your organization. By comparing data sets across events and changes, companies can identify the causes of customer churn. In the healthcare industry, predictive modeling can be used to predict risk of readmission to hospital. Another opportunity that is relevant to almost all industries – seeing patterns that are correlated with employee resignation. 

 How does data mining impact privacy? 

 Although many applications of data mining contribute to a greater good, there are privacy concerns. 

Consent 

It isn’t always clear to consumers when they’ve consented to allow companies to collect data that will be mined. And even when it is, it may not be clear that you don’t have to give consent. 

 

Data breaches 

When you share your data with organizations, you don’t get a lot of say in how it will be stored. If your date isn’t warehoused securely, it may be vulnerable to malicious cybersecurity events.   

 

Data brokering 

If you’re not paying for a service, does that mean you are the product? While you may have consented to allowing an organization to collect your data, you may not be aware that it’s being sold to other organizations. 

 

Data mining offers great potential in helping organizations make better decisions that can improve sales efforts, increase employee retention and even keep people out of the hospital with the right interventions. It also raises several privacy concerns. Organizations conducting data mining need to follow best practices to maintain the trust of consumers. 

 

Contact us today to learn how we can help.