Data Flow

Orchestrate and mine data from disparate enterprise systems together into organised processes.

Build multi-step data processes (e.g. scoring of optimal data, business analytics, ETL processes).

Support for bulk data writing technologies from leading database vendors.

Save money on costly database changes.

Output reports including data and graphs.

Apply batch processing including machine learning algorithms.

Execute arbitrary logic as a part of data transformation processes.

Join heterogeneous data together.

Achieve more interactive websites.

Transform and query data on the fly including dynamic easy to use expressions.

Increase flexibility by achieving automatically changing logic by learning from data.

Query, filter and change data to implement arbitrary business rules.

Achieve better integration between legacy and modern systems.

Trigger processes from the web, scheduling or manual intervention.

Provide simple reporting straight to Excel, Word and PDF.

Perform statistical analysis, and common database operations like aggregations using drag and drop interface.

Take existing data from KnowledgeKube data sources

SQL Providers
Web Services
OData
KnowledgeKube Models

Create derived, in memory data

Understands KnowledgeKube data capture
Create simple/complex queries
Create derived columns using expressions
Create groupings from data or expressions
Use it inside the expression engine Expression engine automatically gets tabular and aggregate values/variables
Push it into another stage in the pipeline

As well as the more traditional data processing tasks, KnowledgeKube data processing can also apply advanced statistical analysis functions and machine learning algorithms.

Advanced Statistical Analysis

Data selection

Group data by influences
Identify trends not directly visible in the data.
Identify trends not considered before.

Generate/approximate values for gaps in data

Output Data
Back into the pipeline.
CSV or XML - Limited visibility.
Graph - Improved visibility (especially when clustering)

Machine Learning Capabilities

Clustering = Abnormal Data Detection

Converting Numerical Data to Categorical Data
Discretization Algorithm.

Adaptive Boosting Classification
Train Data to Improve Predictive Quality.

Classification and Prediction Using Neural Networks
Predict Group Membership, Domains and Properties for New Data or Process for Abnormal Data Detection.

Decision Trees

Build predictive models based on input variables

Visually or explicitly represent decisions in KnowledgeKube models

Map observations to conclusions

KnowledgeKube can access decision trees from the expression parser
Predicate outcomes.
Drive intelligent user interfaces.
Validate input.