• DIKW About DIKW
  • Who are we
  • Philosophy
  • References
  • Boabab Tree
  • DIKW Paradigm
• People Fellow-entrepeneurs
• Services DIKW in practice
  • Advanced Intelligence
  • Competitive Intelligence
  • Data Intelligence
  • Data Logistics
  • Datamining
  • Data Vault
  • Document Management
  • Information Architecture
  • Project Management
  •  Pattern Based Strategy (PBS)
• Solutions unique for DIKW
  • Competitive Intelligence
  • Enterprise Memory
  • Intelligence Factory
  • Social Media Analytics
  • DIKW as a Service
• Training unique for DIKW
• Download Knowledge Base
  • Articles
  • Cases
  • Leaflets
  • Knowledge sessions
  • Books
• Contact DIKW details
  • visit our office

Datamininggenerate value from your data

At DIKW we are "tool agnostic" if it comes to datamining. Open & closed source tools, script or GUI based, all the same to us! We are able to deliver our datamining  compentence on a broad range of datamining tools.

The datamining area can be devided roughly into two parts:

Exploratve Analytics (also known as unsupervised learning methods)

Within the field of explorative analytics there is no predefined target variable, by exploring  and wandering through the data we can discover and exploit possible relationships or structures that were not known before.

Data visualization techniques play an increasingly important role in this field.

Textmining is the field of automated text analysis. In her most simple form it can be to set up a frequency table of most occuring words. In a more advanced setup we parse sentences into their grammatical parts, thus building tagged trees that are more suitable to analyze in an automated way. Typical goals of textmining is to categorize and cluster large document databases.

Applications : Automatic clustering and grouping of larges sets of documents, email or blogs.

Predictive Analytics (also known as supervised learning methods)

Here a clear target value is defined, goal is to build a model that can best predict this outcome. This is always a balancing act between prediction on the training set and performance outside the known cases(also known as robustness). Predictive Analytics aims to deliver a change model that scores customers (or cases), ranking the cases from high to low gives the possibility to select a subset of cases that have a higher change then average of the predicted outcome. Applications : churn/loyalty- , (credit) risk- , cross & up-sell models.

Feel free to have a look at our unique delivery concept for datamining : the "Intelligence Factory" !