Value-Directed Parsing for Ambiguous Date Extraction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing systems struggle with ambiguously specified named entities in unstructured or semi-structured text, particularly dates, which can be expressed in various formats and orders, leading to data formatting challenges.

Innovation Solution

A method for parsing unstructured or semi-structured text to structured data by tokenizing the text, applying value constraints, scoring potential field values, and disambiguating using techniques like learnt orderings, format induction, and neighboring parses to achieve a target value parse.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If multiple formats are used to express named entities, then adaptability is improved, but manufacturing precision deteriorates

Engineering Contradiction:

ImproveadaptabilityVSAvoidprecision

Core Design Contradiction:

Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The system changes the parameter of date representation by accepting multiple format variations (different separators, digit counts, orderings) and transforms them into a standardized internal representation through scoring and disambiguation processes

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent segments the date string into individual components (day, month, year) and evaluates different segmentations separately, assigning scores to each possible interpretation to determine the most likely correct parsing

Inventive Principle:

Principle #1Segmentation

2Ease of operation

If flexible parsing is applied, then ease of operation is improved, but measurement precision deteriorates

Engineering Contradiction:

Improveease of operationVSAvoidprecision

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The system dynamically adjusts its parsing approach by evaluating multiple possible interpretations of the input string and selecting the most appropriate one based on scoring, rather than using a fixed rigid parsing rule

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system uses feedback from reference value parses and learned orderings to improve disambiguation accuracy, comparing candidate parses against known correct examples to refine its interpretations

Inventive Principle:

Principle #23Feedback

3Measurement precision

If disambiguation techniques are applied, then measurement precision is improved, but device complexity increases

Engineering Contradiction:

ImproveprecisionVSAvoidcomplexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary actions by pre-processing the input string to identify potential field values and their possible interpretations before final disambiguation, organizing candidates for efficient evaluation

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces intermediary elements such as reference value parses and learned orderings that mediate between the raw input and final interpretation, providing additional context to resolve ambiguities

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS12596729B2Value-directed parsing for data extraction

Publication Date: 2026.04.07 ZOHO OFFICE SUITE

AI summary

Described are methods and systems for parsing unstructured or semi-structured text to extract named entities, data types defined to include semantic fields. Fields are constrained to sets of potential field values. These sets can overlap, leading to ambiguous parses. For example, the text string “3-4-2023” parsed as a date can yield Mar. 4, 2023 or Apr. 3, 2023. Potential field values in alternative parses are scored and the scores used to select and disambiguate the resultant value parses.