Memory-Based NLP Rule Generation from Natural Language

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional natural language processing techniques are inefficient for dynamically updated policy documents, requiring continuous processing cycles and posing challenges in maintaining rules and constraints due to their inability to adapt effectively to new input parameters.

Innovation Solution

A memory-based natural language processing approach that uses a processor to receive and parse inputs, retrieve similar cases from a repository based on similarity thresholds, adapt these cases to new inputs, generate constraints, and evaluate their accuracy through user feedback, allowing for incremental learning and updating of rules.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional natural language processing techniques are used to process dynamically updated policy documents, then the system can identify and extract policies, but it requires continuous repetition of processing cycles which reduces efficiency and increases maintenance complexity

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidcontinuous processing cycles

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system performs preliminary action by maintaining a repository of previously processed policy cases with extracted rules and constraints. When new policy documents are updated, the system retrieves relevant historical cases and adapts existing rules rather than processing from scratch, significantly reducing processing time and continuous cycle requirements

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses copying by replicating proven rule structures and constraints from historical cases to new policy documents. Instead of re-extracting every rule, the system copies validated rule templates and adapts them to current policy parameters, improving productivity while maintaining accuracy

Inventive Principle:

Principle #26Copying

2Adaptability or versatility

If conventional natural language processing techniques are used to translate policy guidelines into rules, then the system can generate constraints, but it has difficulty adapting to new input parameters which worsens maintenance of existing rules

Engineering Contradiction:

Improveadaptation to new parametersVSAvoidmaintenance complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system implements dynamics by making the rule generation process adaptive and flexible. When new policy parameters or document structures are introduced, the system dynamically retrieves relevant historical cases, adapts existing rules to new parameters, and updates the repository. This dynamic adaptation capability improves versatility while managing maintenance complexity through systematic case-based reasoning

Inventive Principle:

Principle #15Dynamics

3Measurement precision

If conventional natural language processing techniques process large volumes of prospectuses, then the system can identify policies, but the continuous processing cycles increase computational resource consumption

Engineering Contradiction:

Improvepolicy identification accuracyVSAvoidcomputational resource consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system applies extraction by isolating and storing only the essential policy rules and constraints from processed documents in a structured repository. When processing new documents, it extracts only the relevant changes or updates needed rather than re-processing entire documents, significantly reducing computational resource consumption while maintaining identification accuracy

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system uses discarding and recovering by identifying unchanged policy sections from historical cases and reusing them without re-processing. Only modified or new policy sections are processed in full, discarding redundant computational efforts while recovering and reusing validated rule structures, thereby reducing energy consumption

Inventive Principle:

Principle #34Discarding and recovering

Data Source

PatentUS20240320446A1Method and system for memory-based generation of rules from natural language descriptions

Publication Date: 2024.09.26 JPMORGAN CHASE BANK NA

AI summary

A method for translating natural language data into constraints via memory-based processing is disclosed. The method includes receiving inputs via a graphical user interface, each of the inputs including input wording in a natural language format; parsing, by using a model, the inputs to retrieve a case from a case repository, the retrieval including identification of the case based on a similarity value and a predetermined similarity threshold; automatically adapting, by using the model, the retrieved case to the inputs; generating, based on a result of the adapting, constraints that characterize the inputs, the constraints relating to a rule that is mandated by the input; and evaluating the constraints.