Direct Data Mining


Grace systems positions itself on the level of data mining as compared to the classic BI tooling. Instead of laborious data warehouse techniques, data can be analyzed immediately with various modules. Instead of building data couples and reports, business intelligence can be created directly.

Data mining is searching directly for (statistical ) relations in data collections, aimed at constructing profiles for scientific or commercial use. Such a data collection can be formed by registration of various corporate systems in a practical situation (consumer behavior, patient symptoms et cetera) or by comparing and reinterpreting prior scientific research.

Grace system is made up of different modules, supporting various organizational queries relating to the properties of the data.

  • Data Discovery
  • Orphan Discovery
  • Correlation Discovery
  • Correlation Analysis
  • Data Mining
  • Cluster Analysis
  • Rules Engine

Underneath you will find some examples of facilitation and set up of the Grace System, followed by various possible simulations and analyses.

GS_Direct_Data_Mining1

 

GS_Direct_Data_Mining 2


Direct Data Mining

The Data mining module is the answer to the classic BI tooling. Direct analytics are made instead of developing Data warehouses and reports. The mining module makes all possible cross cuts of available data. This substantiates the cohesion and quantity of the available data and is often the basis for further research.

 

GS_Direct_Data_Mining4

 
 
GS_Direct_Data_Mining3


Cluster Analysis

Grace has a number of modules with high-grade statistical algorhythms that look at available data in more intelligent ways than other tooling. For example the ‘Cluster Analysis’ can search for clusters or cohorts of data, i.e. the search for cohesion having to do with customers’ age. In the example below the relatively large cluster of ‘80+’ is striking. The cluster algorhythms will quickly find cohesion in large data collections, such as cost, completion time, age, risk, et cetera.

 
 
GS_Direct_Data_Mining5


Business Rules Engine

Grace’s Business Rules Engine is able to simulate elaborately compiled business rules with various conditions over different corporate data collections. Grace is capable to analyze business questions spread over different administrative corporate systems and to research on which parts of a business rule there is no control. Company rules are put together from rules conditions of the different parts of corporate data. Various stakeholders can check all conditions and check whether they are conformed to the norm. In the example below, customers with opposite expressions on a specific product are being searched for.

 
 


Correlation Analysis

An organization has large quantities of rough data. We demonstrate Grace’s unique ‘Correlation Analysis’ module, with which patterns or exceptions can be identified quickly, looking at data properties from different domains. With the Correlation Analysis aspects can be compared (i.e. product and region in relation to complaints) and correlation or deviation can be found in data patterns.

GS_Collection_big_data3

Every cross cut grants access to the data in the semantic warehouse in order to find explanations for deviation with various stakeholders.

 
 

GS_Collection_big_data6