LLM-Based Activity Name Validation in Process Mining
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Poorly defined activity names in process mining can lead to inaccurate results due to variations in naming conventions, resulting in a large number of activities being identified.
Innovation Solution
Utilizing a large language model to detect flawed activity names by receiving prompts with instructions, validation criteria, and activity names, and determining whether the activity names satisfy the validation criteria, providing explanations and recommendations for improvement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If activity names are manually defined by users, then users can define activity names according to their understanding, but the activity names become poorly defined and inconsistent, resulting in inaccurate process mining results
Solution Approach 1:
The system uses AI to automatically validate and suggest improvements for activity names, allowing the process mining system to self-improve without requiring users to manually define consistent naming conventions. The AI model autonomously identifies naming issues and provides corrections.
Solution Approach 2:
The system provides feedback to users about naming inconsistencies and suggests improved activity names based on validation criteria. This feedback loop helps users understand why certain names are problematic and guides them toward better naming practices.
2Loss of information
If activity names include names, dates, or other components that refer to the same activity, then users can capture detailed information, but the same activity is identified as different activities, resulting in a large number of activities
Solution Approach 1:
The AI validation system automatically detects and consolidates duplicate activities by identifying naming patterns that refer to the same process step, eliminating the need for manual review and consolidation of activity names.
Solution Approach 2:
The system changes the parameter of activity name standardization by applying validation criteria that identify and normalize variations in naming conventions, transforming inconsistent names into standardized forms that represent the same activity.
3Measurement precision
If a large language model is used to validate activity names, then the accuracy and consistency of activity names are improved, but the system complexity increases
Solution Approach 1:
The large language model acts as an intermediary between the raw activity names and the validation criteria. It processes the activity names, compares them against the validation rules, and generates suggestions, serving as an intelligent mediator that simplifies the overall validation process.
Solution Approach 2:
The patent replaces manual mechanical validation processes with an AI-based system that automatically validates activity names according to validation criteria, eliminating the need for human reviewers to manually check each activity name for consistency and accuracy.
Data Source
AI summary
Systems and methods for determining whether activity names of a process satisfy a validation criteria are provided. One or more prompts defining 1) instructions, 2) activity name validation criteria, and 3) activity names of a process are received. It is determined whether the activity names satisfy the validation criteria using a large language model based on the instructions. Results of the determining are output.