API Platform Data Linkage via Predictive Request Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data linkage systems face inefficiencies in processing and responding to API requests due to unpredictable processing times and data amounts, leading to suboptimal resource allocation and performance.
Innovation Solution
A data linkage system that includes a data collection system, a data storage system, and an API platform with servers that classify API requests based on predicted processing time and response data amounts, using machine learning to determine appropriate resource allocation and processing paths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If API requests are processed without classification, then the system structure remains simple, but processing efficiency deteriorates due to unpredictable processing times and data amounts
Solution Approach 1:
The API platform segments incoming requests into different queues based on predicted processing time levels (first, second, third queues) and predicted response data amount levels (first, second queues). This segmentation allows each queue to be handled by appropriately sized processing resources, improving overall processing efficiency while maintaining manageable system complexity through structured organization.
Solution Approach 2:
The system changes the parameter of request handling by introducing classification based on predicted processing time and data amount. Requests are transformed from a single undifferentiated stream into categorized groups with different handling characteristics, enabling optimized resource allocation and processing paths for each category.
2Productivity
If resources are allocated uniformly to all API requests, then resource allocation is simple, but system performance deteriorates due to unpredictable processing times and data amounts
Solution Approach 1:
The system applies local quality by allocating different resource levels to different request queues based on their specific needs. First queues (shorter processing time, smaller data amount) receive standard resources, while third queues (longer processing time, larger data amount) receive escalated resources. This localized resource quality optimization improves system performance without requiring complex dynamic reallocation.
3Productivity
If the system processes all API requests through a single path, then the processing path is simple, but responsiveness deteriorates due to varying processing requirements
Solution Approach 1:
The processing path is segmented into multiple parallel paths corresponding to different queue combinations (first/second/third queues for processing time, first/second queues for data amount). Each path is optimized for its specific request type, enabling the system to respond more effectively to varying processing requirements while maintaining clear, manageable path structures.
Data Source
AI summary
A data linkage system includes an API platform that provides an API for acquiring data, which is based on data stored in a data storage system for storing data collected by a data collection system and held by an information system, from the data storage system. The API platform includes servers that execute processing to acquire the data from the data storage system in response to a request for use of the API for each classification based on at least one of a predicted processing time level, which indicates duration of a predicted time for the processing to acquire the data in response to the request for the use of the API, and a predicted response data amount level, which indicates a predicted data amount responding to the request.