Semantic Action Graphs for Accurate RPA Task Recommendations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotic process automation (RPA) technologies fail to efficiently automate repetitive tasks due to inefficiencies in task mining and process determination, leading to reduced productivity and increased user workload.
Innovation Solution
The implementation of semantic action graphs using task mining data, normalization techniques, classification algorithms, clustering algorithms, reinforcement learning, and supervised learning to generate actionable insights for RPA, enabling automated task recommendations and workflow optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional task mining methods are used in RPA, then the system can identify repetitive tasks, but the efficiency and accuracy of task automation determination is insufficient
Solution Approach 1:
The patent transforms raw task mining data into normalized data by applying normalization techniques that convert diverse task parameters into a unified scale. This parameter transformation enables accurate comparison and classification of tasks across different applications and users, resolving the contradiction between processing efficiency and classification precision.
Solution Approach 2:
The patent introduces semantic action graphs as an intermediary structure that connects normalized task data with automation recommendations. These graphs serve as a mediator between raw task mining data and final automation decisions, enabling both efficient processing and accurate determination by decomposing complex tasks into actionable semantic units.
2Reliability
If manual task analysis is performed for RPA implementation, then detailed process understanding is achieved, but user workload increases and productivity decreases
Solution Approach 1:
The patent implements self-service automation determination by enabling the system to automatically analyze task mining data, generate semantic action graphs, and provide automation recommendations without requiring manual intervention. The system serves itself by using its own collected data to identify automation opportunities, eliminating the need for separate manual analysis phases and reducing user workload while maintaining high accuracy.
Solution Approach 2:
The patent performs preliminary normalization and semantic graph generation on task mining data as it is collected, preparing the data structure in advance for automation determination. This preliminary processing ensures that when automation recommendations are needed, the analysis can be performed quickly on already-processed data, reducing the time loss associated with on-demand manual analysis.
3Adaptability or versatility
If comprehensive task mining data is collected from multiple user systems, then better automation patterns are identified, but data processing complexity increases
Solution Approach 1:
The patent segments comprehensive task mining data into normalized units that can be independently processed and then recombined. By dividing the large-scale multi-system data into standardized semantic action components, the system achieves better pattern recognition across diverse sources while keeping the processing complexity manageable through modular normalization techniques.
Data Source
AI summary
Action and/or process determination and recommendations for Robotic Process Automation (RPA) using semantic action graphs is disclosed. Semantic action graphs are graphs that store individual actions, and potentially graphical elements and/or text associated with the actions, as nodes, as well as the relationships between nodes as edges. Metadata to develop the semantic action graphs may be derived from task mining applications that can monitor the interactions of users with computing systems, workforce intelligence, etc. The semantic action graphs may be for a user, an organization, an industry, product-wide, etc. At their lowest level of granularity, the recommendations may be for mouse clicks, key presses, Application Programming Interface (API) calls, system events, etc. At higher levels of granularity, the recommendations may be for opening an order, creating a lead, approving a work item, etc.