Automated Pattern Template Generation for Bulk Text Messages
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for extracting information from bulk text messages, such as SMS, are inefficient and require extensive human effort, as they rely on manual regular expressions (regex) that are time-consuming and limited by human resources, and often result in narrow focus and performance issues when processing multiple regexes.
Innovation Solution
An automated system that processes messages to identify string and numeric identifiers, replaces them with labels to generate candidate templates, and merges these templates based on pre-determined rules to create pattern templates, reducing the need for manual intervention and improving efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual regular expressions are used for information extraction, then extraction accuracy can be maintained, but extensive human effort and time are required
Solution Approach 1:
The system enables automated template generation that performs self-service by automatically learning patterns from messages and generating extraction templates without requiring manual regex creation. The automated system processes messages, identifies patterns, and generates templates independently, eliminating the need for continuous manual intervention while maintaining extraction accuracy.
Solution Approach 2:
The patent replaces the manual mechanical process of writing and maintaining regular expressions with an automated computational system. The system uses algorithms to automatically learn patterns from messages and generate extraction templates, substituting human manual work with automated computational processes that are faster and more scalable.
2Adaptability or versatility
If multiple regular expressions are written for same features, then coverage is improved, but system complexity and processing overhead increase
Solution Approach 1:
The system merges multiple extraction templates into unified patterns by identifying common structures across different messages. Instead of maintaining separate regexes for similar features, the system combines them into consolidated templates that cover multiple cases, reducing system complexity while maintaining or improving message coverage.
Solution Approach 2:
The generated templates are designed to be universal and multi-functional, capable of extracting multiple features from different message types using a single template. This universality reduces the number of specialized templates needed, thereby reducing system complexity while maintaining broad coverage.
3Manufacturing precision
If manual regex writing is performed, then extraction precision can be controlled, but productivity is limited by human resources
Solution Approach 1:
The automated system performs self-service by automatically learning from messages and generating extraction templates without human intervention. This enables the system to handle large volumes of messages simultaneously, dramatically increasing extraction throughput while maintaining precision through algorithmic consistency.
Solution Approach 2:
The system performs preliminary learning from message samples to pre-generate extraction templates before actual extraction operations. This preliminary action enables the system to be ready for high-volume processing, increasing productivity while maintaining precision through pre-validated templates.
Data Source
AI summary
An automated pattern template generation system is provided. The system includes a memory having computer-readable instructions stored therein and a processor. The processor is configured to access a plurality of messages and process each of the plurality of messages to identify presence of string identifiers and numeric identifiers present in each of the plurality of messages. The processor is further configured to replace an identified numeric identifier in each message by a numeric label to generate a candidate template corresponding to the message. Furthermore, the processor is configured to merge the generated candidate templates based on pre-determined merging rules to replace one or more identified string identifiers with pre-determined string labels to generate the pattern templates. Each of the generated pattern templates includes numeric labels, string labels, or combinations thereof.