Analytics in Real-Time with Real Time Platform (RTTK)
The RTTK platform uses Complex Event Processing (CEP) to allow data-dependent processing and analysis of captured data in the form of Fundamental Data Artifacts (FDA). Data can be filtered, down-sampled, combined, or synthesized as triggered by “Complex Events.”
Capturing an FDA is considered an “Event” within RTTK. “Complex Events” are a higher level event that is inferred from occurrences of specified lower level events. For example, consider three Events: (1) church bells ring; (2) a couple in formal apparel emerges from the church; and (3) rice flies in the air. The occurrence of these three Events at the same place and time would lead one to infer that there is a Complex Event, wedding. Complex Event Processing is the performance of operations on Complex Events, including reading, creating, transforming, and making inferences from them.
- Event – Something that happens at a point in time. In the context of RTTK operation, FDA capture can be viewed as a simple event.
- Complex Event – A higher level event that is inferred from the occurrences of lower level events.
Events: Church bells ring, couple emerges and rice flies in air.
Complex Event: Wedding
- Complex Event Processing – Performing operations on complex events, including reading, creating, transforming, and making inferences on them. Inference Rules IF THEN …Representation: Object Management Group (OMG) open standard “Production Rule Representation (PRR)”
In the RTTK platform, FDAs are detected as they are captured. The logic for processing them is expressed as declarative “inference rules” associated with each FDA. An inference rule has the form, “IF <condition> THEN <action1> … <actionN>”. The design for rules in RTTK conforms to the open standard “Production Rule Representation (PRR)”. This adherence to an open standard maximizes the potential opportunities for import/export of rules from other tools or allows a project to replace the FishEye implementation with a third party tool.
Example CEP Rule
CEP evaluates rules and, when the condition evaluates to “true”, executes each of the actions. Rules are defined in an XML file, which is read at startup and stored efficiently for real-time use. Action expressions can create new FDAs, retract pending FDAs, or perform any combination of assert/retract actions. The assert/retract capability can be used to filter large, unwieldy FDAs and replace them with a set of smaller FDAs.
To learn more or for a free system assessment, contact us.