Dynamic Metric Selection Using Registration Structures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current products for programmatic analysis of data repositories are inflexible and require strict knowledge of input data sets, using statically defined metric calculation methods that lack awareness of applicable metrics without hard-coded associations, making them unsuitable for diverse data types and remote storage environments.
Innovation Solution
A system that dynamically selects and executes metric applications against a data set by using a metric registration structure to map record key data to applicable metric applications, allowing for dynamic definition and execution of metrics, with support for arbitrary code generation and execution, and output storage of calculation results and intermediate records.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If statically defined metric calculation methods are used, then the system structure is simple and easy to implement, but the system lacks flexibility and cannot adapt to diverse data types
Solution Approach 1:
The patent implements dynamic metric selection by allowing the system to automatically determine which metrics to execute based on the actual input data characteristics, rather than using fixed static definitions. The metric registration structure enables runtime registration and selection of metric applications, making the system adaptable to diverse data types while maintaining a manageable structure through automated classification.
Solution Approach 2:
The system changes the parameter of metric selection from static pre-definition to dynamic determination based on data characteristics. By using record key data and data type information to automatically select appropriate metrics, the system achieves versatility without requiring complex manual configuration for each data type.
2Productivity
If manual identification of applicable metrics is performed, then the system can ensure accurate metric selection, but the process requires significant manual intervention and time
Solution Approach 1:
The system performs self-service by automatically identifying and selecting applicable metrics based on the input data's record key data and data types. The metric registration structure enables the system to autonomously determine which metrics to execute without requiring manual intervention, thereby improving productivity while maintaining accurate metric selection through automated data-driven decisions.
Solution Approach 2:
The system uses feedback from the input data characteristics (record key data, data types) to automatically select appropriate metrics. The metric registration structure stores mappings between data characteristics and applicable metrics, allowing the system to receive feedback from the data and automatically adjust metric selection, eliminating manual intervention while ensuring accuracy.
3Reliability
If all metrics are executed against all data sets, then comprehensive analysis is achieved, but the system performs unnecessary calculations and wastes computational resources
Solution Approach 1:
The patent applies partial action by executing only the subset of metrics that are applicable to the specific input data, rather than running all metrics universally. The metric registration structure enables the system to identify and execute only the relevant metrics based on data characteristics, ensuring complete coverage of applicable metrics while avoiding unnecessary calculations and reducing computational resource waste.
Solution Approach 2:
The system segments the metric execution process by dividing all available metrics into applicable and inapplicable groups based on the input data's record key data and data types. The metric registration structure facilitates this segmentation by storing associations between data characteristics and specific metrics, allowing the system to execute only the relevant segment of metrics, thereby maintaining reliability while reducing energy loss.
4Adaptability or versatility
If the system requires strict knowledge of input data sets, then metric calculations can be accurately performed, but the system cannot handle remote storage and diverse data types
Solution Approach 1:
The patent implements universality by creating a metric registration structure that can handle diverse data types and remote storage environments through a unified interface. The system uses record key data and data type information as universal identifiers to select appropriate metrics, allowing accurate metric calculations across different data sources and formats without requiring strict knowledge of each specific input data set's structure.
Data Source
AI summary
A method, apparatus and computer program product are provided for generation, selection, and execution of metric applications. An example of the method includes receiving, via record access circuitry, the set of record data from a record repository, determining, by a processor and from the set of record data, record key data and one or more record data tables for the set of record data, selecting, by metric management circuitry, one or more metric applications based on a mapping performed between record key data and a metric registration structure, wherein the metric registration structure comprises metric application metadata indicating data types required by each of a plurality of metric applications, executing, by a processor, each of the selected one or more metric applications, determining an output of each of the executed selected one or more metric applications, and storing the output in a memory.