Temporal Knowledgebase Memory Management via Unique Tuple Tickets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in maintaining unique data values in a knowledgebase while optimizing memory usage and ensuring real-time decision-making based on input events, as they often suffer from memory overflow due to redundant data storage.
Innovation Solution
A working memory module with a tuple store sub-module, data store sub-module, threshold store sub-module, and state variable store sub-module is implemented, utilizing a weak hash map to ensure unique data values are maintained and automatically purged when no longer referenced, along with a method to determine the real-time significance of events within a defined time frame for triggering actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all input events are stored in the knowledgebase for real-time decision-making, then the system can ensure complete data availability for analysis, but memory overflow occurs due to redundant data storage
Solution Approach 1:
The patent extracts only the essential and unique features from input events using a weak hash map data structure. Instead of storing complete event objects, the system extracts unique data values and maintains only those that are currently referenced by active rules, removing redundant copies and reducing memory consumption while preserving data availability for decision-making
Solution Approach 2:
The system implements automatic purging of data values from the knowledgebase when they are no longer referenced by any active rules. The weak hash map structure allows the system to discard redundant data copies and recover memory space, while maintaining references to unique values that are still needed for ongoing decision-making processes
2Loss of information
If a traditional data structure is used to store all events, then data completeness is maintained, but system performance degrades due to memory management overhead
Solution Approach 1:
The patent uses a weak hash map that creates lightweight references to data values rather than storing complete copies of all event objects. The hash map stores only the essential unique values needed for pattern matching, reducing memory management overhead and improving system performance while maintaining data completeness through reference integrity
Solution Approach 2:
The system changes the data structure parameters from storing complete event objects to storing only unique data values with weak references. This parameter change reduces the memory footprint and management overhead, thereby improving system performance while the reference mechanism ensures data completeness is maintained for active correlations
Data Source
AI summary
Systems and methods are described to address the problem of maintaining only unique data values accessed through name-value tuple sets. A knowledgebase represents data in tuple sets where all the data across the tuples comprise unique instances for memory conservation and system performance purposes. The knowledgebase stores and maintains unique values in a data store structure and represents the values as tuples through the tuple store structure. Modules access tuples and their values through handles called tuple tickets. The knowledge base defines a set of valid data types that can be maintained, and all data types are accessible via a common interface. In addition, memory management is optimized using a system that frees objects that are no longer referenced.