Process Mining Workflow Orchestration Using Multimodal Event Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional orchestration platforms fail to automatically orchestrate entire business processes across multiple domains, focusing either on DevOps or containerization, and lack the ability to handle complex, multimodal business process data effectively.
Innovation Solution
A system utilizing unsupervised learning to extract and correlate multimodal event data, generating process models that automatically orchestrate workflows by associating event instances with vector representations and applying LSTM neural networks for pattern recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional orchestration platforms focus on DevOps or containerization-specific operations, then they can effectively manage those specific domains, but they fail to automatically orchestrate entire business processes across multiple domains
Solution Approach 1:
The patent creates a universal orchestration platform that can handle multiple domains (DevOps, business processes, containerization) through a single system. The process mining engine extracts events from multiple sources and the generated workflow models can orchestrate跨域 business processes, making the platform multi-functional rather than domain-specific
Solution Approach 2:
The patent introduces process mining technology as an intermediary layer that bridges different domains. By extracting events from various sources (logs, databases, APIs) and generating unified workflow models, the system mediates between domain-specific operations and cross-domain orchestration, enabling automated business process management across multiple domains
2Loss of information
If the system processes complex multimodal business process data from multiple sources, then it can achieve comprehensive process mining, but the difficulty of detecting and measuring patterns increases
Solution Approach 1:
The patent segments the complex multimodal data processing into distinct components: event extraction from multiple sources, event correlation through process mining, and workflow generation. The system divides the large-scale data into event instances that can be individually processed and correlated, making the overall pattern recognition manageable despite data complexity
Solution Approach 2:
The patent replaces manual pattern detection with automated process mining algorithms and machine learning models. The system uses computational methods to automatically extract patterns from multimodal data, substituting human analysis with algorithmic processing that can handle the complexity of cross-domain business process data
3Measurement precision
If manual monitoring and analysis of business processes is performed, then detailed process understanding can be achieved, but time consumption and inefficiency increase
Solution Approach 1:
The patent implements self-service process mining where the system automatically extracts events from data sources, correlates them to discover process patterns, and generates workflow models without human intervention. The automated orchestration engine then executes the discovered processes, enabling the system to serve itself in monitoring and optimizing business processes
Solution Approach 2:
The patent establishes feedback loops where process mining continuously monitors business process execution, discovers patterns and bottlenecks, and automatically adjusts workflow orchestration. The system uses feedback from event data to refine process models and improve orchestration decisions, enabling continuous process optimization without manual analysis
Data Source
AI summary
A system for discovering business processes using unsupervised learning is configured to: (a) receive multimodal event data from a plurality of sources, the multimodal event data including a plurality of event instances; (b) associate the multimodal event data with a vector representation, such that the plurality of event instances is represented as a plurality of event vectors; (c) correlate the plurality of event vectors using unsupervised learning to identify one or more processes; and (d) generate a process model script for the one or more processes. A method for automated orchestration of a workflow is also disclosed.