Distribution Trees for Form Field Confidence Modeling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

There is a need to accurately determine the accuracy of input form field data in electronic documents to ensure correct results in computerized services, such as tax preparation, where data is often manually entered, retrieved using OCR, or pulled via APIs, and errors can lead to incorrect calculations.

Innovation Solution

A multi-part confidence model using inter-field correlations, trained distribution trees, featurization logic, and a classifier model to assess the accuracy of input data by analyzing the patterns and distributions of values across multiple fields in a document, providing a confidence score for each field's input value.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If form field data is manually entered or retrieved using OCR/API, then data input efficiency is improved, but data accuracy deteriorates

Engineering Contradiction:

Improvedata input efficiencyVSAvoiddata accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system implements feedback by using distribution trees to predict expected values for form fields based on inter-field correlations. The predicted values are compared with actual input values, and confidence scores are generated to indicate whether the input data is likely correct. This feedback mechanism allows the system to automatically detect potential errors without requiring manual verification of each field.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The distribution tree acts as an intermediary between the input data and the accuracy assessment. It uses features from multiple other fields in the document to predict the expected value of a target field, serving as a mediator that translates complex inter-field relationships into a confidence score for data accuracy.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Ease of operation

If simple data input methods are used, then ease of operation is improved, but reliability deteriorates

Engineering Contradiction:

Improveease of data inputVSAvoidreliability of input data

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The system performs self-service by automatically assessing the accuracy of user-input data without requiring user intervention. The distribution trees and classifier model autonomously analyze inter-field correlations, generate predictions, and provide confidence scores, allowing the system to self-verify data quality while users simply input data using convenient methods.

Inventive Principle:

Principle #25Self-service

3Device complexity

If traditional single-field validation is used, then device complexity is minimized, but measurement precision deteriorates

Engineering Contradiction:

Improvevalidation system complexityVSAvoidaccuracy assessment precision

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The validation system is segmented into distinct functional components: distribution trees for prediction, feature extraction logic for data processing, and a classifier model for confidence scoring. Each component handles a specific aspect of the validation process, allowing the system to achieve high measurement precision through specialized sub-components while maintaining manageable overall complexity through modular architecture.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11430237B1Systems and methods utilizing distribution trees for confidence modeling and use of form field data

Publication Date: 2022.08.30 INTUIT INC

AI summary

Systems and methods that may be used to determine that input form field data is accurate or not, and associate a level of confidence with that determination. The systems and methods may use a multi part confidence model that uses inter-field correlation to tie the correctness of a particular field to the pattern of values seen in other fields of the document the field data is input from.