In-Memory Temporal Processing for Scalable Event Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Business Intelligence (BI) systems face challenges in scalable and affordable complex event processing, especially with real-time processing of large data sets, due to high computational costs and interpretation issues from heterogeneous event formats and domain expertise encoding.
Innovation Solution
An in-memory temporal processing system is introduced, utilizing an in-memory database and semantic event modeling with RDF triples, combined with a temporal rules engine to process and analyze event streams, enabling efficient real-time processing and maintaining expressiveness and scalability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex event processing is performed at the application level, then relevant information can be filtered and analyzed, but computational costs and processing time increase significantly
Solution Approach 1:
The patent introduces a new dimension by implementing event processing rules at the database engine level rather than at the application level. This dimensional shift in where processing occurs allows the system to leverage database optimization capabilities, reducing computational overhead while maintaining analysis accuracy. The database engine becomes an active participant in event processing, transforming the traditional application-centric architecture.
Solution Approach 2:
The patent introduces an intermediary layer between the event stream and the application layer - specifically, a database engine with embedded event processing capabilities. This intermediary handles the computationally intensive filtering and analysis of events, preventing these operations from burdening the application layer. The database engine acts as a mediator that translates raw event streams into processed information efficiently.
2Adaptability or versatility
If heterogeneous event formats and domain expertise encoding are used, then flexibility in representing different data types is achieved, but interpretation issues arise across systems
Solution Approach 1:
The patent applies homogeneity by standardizing event representation through a unified schema within the database engine. While the system accepts heterogeneous event formats from different sources, it transforms them into a homogeneous internal representation that ensures consistent interpretation across all systems. This standardization layer maintains flexibility in data sources while guaranteeing reliability in processing.
Solution Approach 2:
The database engine serves as an intermediary that mediates between heterogeneous event formats and the processing logic. It provides a translation layer that converts diverse input formats into a standardized internal representation, ensuring that domain expertise encoding is interpreted consistently regardless of the source system's native format.
3Loss of time
If real-time processing of large data sets is implemented, then timely business insights are provided, but system performance and scalability are challenged
Solution Approach 1:
The patent implements self-service by enabling the database engine to autonomously process events using embedded rules and procedures. The system performs real-time filtering, aggregation, and analysis without requiring external application intervention for each event. This self-service capability at the database level reduces latency and maintains high throughput even with large data sets.
Solution Approach 2:
The patent applies preliminary action by pre-compiling event processing rules and procedures into the database engine before runtime. Domain expertise is encoded into optimized execution plans in advance, allowing the system to process events in real-time without the overhead of interpreting complex logic during event processing. This pre-preparation enables both speed and scalability.
Data Source
AI summary
A method, medium, and system to receive an event stream, the event stream including a plurality of events, the events being semantically modeled; receive domain insights specifying a relationship between two events, the domain insights being semantically modeled and defined by a specified time limit and a comparison of event attributes using the specified time limit with a logical operator; retrieve stored representations of events referenced in the received domain insights; process the event stream, the received domain insights, and the retrieved stored events to produce a temporal processing result; and store the temporal processing result.