Job Sequence Analysis via Time Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for analyzing the sequence relation between batch jobs are inefficient, especially in large-scale systems, as they face increased computational complexity with a growing number of tasks, making it difficult to obtain results within a reasonable time, particularly due to the assumption of processes executed by humans with only a few tasks.
Innovation Solution
A method that classifies jobs based on time segments, generates data indicating execution sequence relations within and across these segments using end time points, and analyzes these relations to prevent exponential processing time growth, allowing for efficient job sequence analysis by partitioning and linking execution sequences across segments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If process mining techniques (e.g., α-algorithm, genetic algorithm) are used to analyze sequence relations between batch jobs, then measurement precision of job sequence relations is improved, but computational complexity increases exponentially with the number of jobs
Solution Approach 1:
The patent divides the time period into multiple time segments and partitions the batch jobs into different groups based on their execution time periods. By analyzing sequence relations within each partition separately and then combining the results, the computational complexity is reduced from exponential to linear with respect to the number of jobs, while still maintaining accurate sequence relation identification across the entire job set.
2Reliability
If process mining techniques are applied to large numbers of batch jobs, then completeness of sequence analysis is improved, but processing time becomes unreasonably long
Solution Approach 1:
The patent segments the batch job set into multiple partitions based on time periods and analyzes each partition independently. This segmentation allows the system to maintain complete analysis coverage while reducing processing time by avoiding the exponential computational burden of analyzing all jobs simultaneously. The results from individual partitions are then integrated to provide complete sequence relation information.
3Measurement precision
If traditional process mining algorithms are used without partitioning, then accuracy of sequence relation detection is improved, but productivity of the analysis system deteriorates due to exponential time complexity
Solution Approach 1:
The patent implements time-based segmentation of batch jobs into multiple partitions, allowing the system to process and analyze job sequence relations in smaller, manageable groups. This segmentation maintains the accuracy of sequence relation detection by applying mining algorithms to each partition while dramatically improving productivity by reducing the computational complexity from exponential to linear, enabling the analysis of large numbers of jobs within reasonable time frames.
Data Source
AI summary
A job analyzing method includes classifying jobs in log data in accordance with a time segment to which an end time of each of the jobs belongs; generating, for first jobs included in a first time segment, first data indicating an execution sequence relation between the first jobs based on end time of the jobs, and generating, for second jobs included in a second time segment succeeding the first time segment, second data indicating an execution sequence relation between the second jobs based on end time of the second jobs; and analyzing an execution sequence relation between the first and second jobs based on the end time of the first jobs and the end time of the second jobs, and generating data indicating the execution sequence relation between the first and second jobs across the first and second time segments.