Nonnegative Matrix Factorization for Business-Process Resource Usage Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods face challenges in analyzing time-series data indicating temporal variation of resource usage states by business-process, as data integration across multiple processes complicates individual business-process analysis, making it difficult to separate and identify component data specific to each process.
Innovation Solution
The approach involves generating an operation-data matrix from time-series data, performing nonnegative matrix factorization to extract basis vectors representing component values for resources, and using these to output information about resource usage states specific to each business-process, leveraging periodic variation tendencies and processing amounts to separate data by business-process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If time-series data from multiple processes is integrated for analysis, then comprehensive resource usage information is obtained, but it becomes difficult to separate and identify component data specific to each individual process
Solution Approach 1:
The patent applies segmentation by decomposing the integrated time-series data into process-specific components through nonnegative matrix factorization. The operation data matrix representing multiple processes is segmented into basis vectors and weight vectors, where each basis vector corresponds to a specific process's resource usage pattern. This allows the system to maintain comprehensive resource usage information while simultaneously identifying and separating individual process contributions.
Solution Approach 2:
The patent introduces nonnegative matrix factorization as an intermediary mathematical technique that bridges the gap between integrated time-series data and process-specific component data. This intermediary method transforms the mixed data into a structured format with basis vectors representing individual processes and weight vectors representing temporal variations, enabling both comprehensive analysis and process-specific identification.
2Ease of operation
If nonnegative matrix factorization is performed on operation data matrix to extract process-specific components, then individual business-process analysis is enabled, but computational complexity increases
Solution Approach 1:
The patent applies parameter changes by transforming the operation data matrix into a different parameter space through nonnegative matrix factorization. Instead of analyzing the original high-dimensional time-series data directly, the system transforms it into basis vectors and weight vectors with fewer dimensions that capture the essential process-specific patterns. This parameter transformation simplifies subsequent business-process analysis while the computational complexity is managed through efficient matrix factorization algorithms.
3Measurement precision
If basis vectors representing component values are extracted for each resource, then resource usage trends by process can be identified, but data processing time increases
Solution Approach 1:
The patent applies preliminary action by pre-computing the nonnegative matrix factorization of the operation data matrix to extract basis vectors representing resource usage patterns for each process. This preliminary extraction of component data structures enables efficient subsequent queries and analyses of resource usage trends without repeatedly processing the entire time-series dataset. The computational burden is performed once upfront, reducing processing time for individual process analyses.
Data Source
AI summary
Time-series data indicating a temporal variation of an index, which indicates a usage state of each of resources that are used by multiple processes, is acquired, and an operation-data matrix including vectors is generated based on the time-series data such that each of the vectors indicates the time-series data at a predetermined time interval and includes as an element the index indicating the usage state of one of the resources at the predetermined time interval. A basis matrix including a predetermined number of basis vectors is generated by performing nonnegative matrix factorization on the operation-data matrix. Component values, which respectively correspond to the resources, indicated by each of the predetermined number of the basis vectors are extracted, and information on the extracted component values is output as usage states of the resources that are used by each of the multiple processes.