Real-Time Event Data Analysis Using Contextual Knowledge Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AI and machine learning systems lack the capability to assign context to data from diverse environments, leading to inefficient data labeling and scalability issues, and they often require replicating analytics environments for multi-tenancy, posing challenges in real-time data analysis and security, especially in regulated industries.
Innovation Solution
A system that classifies and assigns unique credentials to client devices, applies restrictions for data streaming, and uses AI-driven machine learning to analyze and standardize event data objects, generating insights in real-time by creating knowledge graphs and dependency maps, enabling contextual data processing and multi-tenancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing AI and machine learning systems are used to analyze data from diverse environments, then data analysis can be performed, but the systems lack the capability to assign context to data, leading to inefficient data labeling and scalability issues
Solution Approach 1:
The patent applies preliminary action by pre-defining context templates and data structure schemas before data ingestion. The system establishes contextual frameworks, event type classifications, and relationship models in advance, allowing data to be automatically contextualized upon arrival without requiring complex real-time analysis or manual labeling, thereby improving efficiency while maintaining manageable system complexity
Solution Approach 2:
The system transforms uncontextualized raw data into contextualized structured data by changing key parameters: adding context identifiers, assigning event types, establishing relationships, and structuring according to pre-defined schemas. This parameter transformation enables the data to become meaningful and analyzable without requiring the system to invent complex contextualization logic for each data point
2Reliability
If existing AI systems are implemented within closed environments with physical and logical separation for multi-tenancy, then data security and privacy can be ensured, but challenges arise when extracting real-time results from closed systems
Solution Approach 1:
The patent introduces an intermediary layer (the event data object structure with standardized context fields) that sits between the closed multi-tenant systems and the AI analysis engine. This intermediary allows data to be securely ingested from isolated sources, contextualized in a standardized format, and then made available for real-time analysis without breaking the security boundaries of the closed environments
Solution Approach 2:
The system implements a universal event data object structure that can accommodate data from multiple different closed systems and tenancy configurations. This universal format enables the same AI engine to process data from diverse sources while maintaining the security and isolation requirements of each individual system, making real-time extraction feasible across multiple closed environments
3Adaptability or versatility
If different data collection mechanisms and analytical engines are built for each specific use case, then tailored analysis can be achieved, but the complexity of maintaining multiple separate systems increases
Solution Approach 1:
The patent segments the data processing function into two independent parts: a universal data ingestion and contextualization layer that handles all data sources, and specialized analytical engines that can be selectively applied. This segmentation allows the system to maintain a single, simple data collection mechanism while offering multiple tailored analysis capabilities through modular engines that can be activated based on specific use cases, reducing overall maintenance complexity
Solution Approach 2:
The system creates a universal event data object structure and contextualization framework that serves all use cases, while allowing multiple analytical engines to operate on the same standardized data. This multi-functional design enables a single platform to deliver tailored analysis for different scenarios without requiring separate data collection and processing mechanisms for each use case
Data Source
AI summary
System and method for analyzing event data objects in real-time in a computing environment are disclosed. The system receives application data from various endpoints. Unique credentials are assigned to client and sub-client devices for each application, allowing restrictions to be applied to streaming of application data associated with specific identifiers. The received event data objects are stored in a database, following predefined formats, and applying endpoint-specific restrictions. Metadata is assigned to each stored event data object, and corresponding output data is also stored in a database based on assigned metadata. The validity parameters of output data are analyzed using ML techniques and data standardization. By correlating event data objects based on validity parameters, a knowledge graph is generated. Real-time analysis of downstream data is performed based on this weightage, leading to a generation of insights, ML-based insights, and AI-based insights.