Machine Learning Model for Automated Entity Field Correction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The challenge lies in processing prescription fill requests received via fax, where the lack of standardized electronic field labels and varied formats makes it difficult to automatically identify necessary data elements for processing, leading to inefficiencies in filling prescriptions.

Innovation Solution

A computer system equipped with a machine learning model that scans database entities, generates structured scan data, identifies missing field values, and uses similarity metrics like Levenshtein and Jaro-Winkler distances to predict and validate missing field values, enabling automated processing of prescription fill requests.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a single blueprint is used to specify exact location of data fields, then processing consistency is improved, but adaptability to different fax formats deteriorates

Engineering Contradiction:

Improveprocessing consistencyVSAvoidadaptability to different fax formats

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The system dynamically adapts its field location expectations based on the actual document format detected. Instead of using a fixed blueprint, the system learns and adjusts to different fax layouts through machine learning models, allowing it to maintain processing consistency across varied formats.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes its processing parameters (field location expectations, extraction methods) based on the detected document type and format. This allows the same processing system to handle multiple fax formats by adjusting its internal parameters rather than requiring different systems for different formats.

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If manual verification of missing field values is performed, then accuracy is improved, but processing time deteriorates

Engineering Contradiction:

Improveaccuracy of field valuesVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs self-verification of predicted field values by automatically comparing predicted values against the scanned document content using similarity metrics (Levenshtein distance, Jaro-Winkler distance). This self-service verification eliminates the need for manual verification while maintaining high accuracy.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual verification processes with automated machine learning-based verification systems. The mechanical process of human review is substituted with computational methods that use similarity metrics to verify field values, dramatically reducing processing time while maintaining or improving accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Productivity

If machine learning models are used to predict missing field values, then processing speed is improved, but complexity of the system deteriorates

Engineering Contradiction:

Improveprocessing speedVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The machine learning models serve multiple functions: predicting missing field values, verifying accuracy through similarity metrics, and adapting to different document formats. This multi-functionality consolidates what would otherwise require multiple separate systems into a single unified model, managing complexity while improving productivity.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The machine learning models act as an intermediary layer between the scanned document and the final processed data. This intermediary handles the complexity of pattern recognition and value prediction, allowing the rest of the system to remain relatively simple while achieving high processing speeds.

Inventive Principle:

Principle #24Intermediary (Mediator)

4Reliability

If similarity metrics are used to validate predicted values, then reliability of automated processing is improved, but computational resources required deteriorates

Engineering Contradiction:

Improvereliability of automated processingVSAvoidcomputational resources

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The system applies similarity metrics selectively rather than universally - only when and where needed to validate predicted field values. This partial application of computational resources maintains high reliability for critical validations while reducing overall computational consumption compared to applying metrics to all data elements.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS20230162831A1Machine learning models for automated entity field correction

Publication Date: 2023.05.25 EXPRESS SCRIPTS STRATEGIC DEVELOPMENT INC

AI summary

A method for automated entity field correction includes scanning a database entity to generate structured scan data, generating a feature vector input according to the structured scan data, and processing, by a machine learning model, the feature vector input to generate an entity field output including multiple identified entity fields and values of the identified entity fields. In response to determining that the entity field output includes at least one missing field value, the method includes accessing a record database to identify a predicted value for the missing field value, comparing a predicted name string of the predicted value to a scanned name string of the structured scan data to determine at least one of a Levenshtein distance and a Jaro-Winkler distance, and transmitting the database entity to a prescription fill processing module for automated processing of a prescription fill specified by the database entity.