Logical Form Parsing With Ambiguous Date Encoding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing natural language to logical form (NL2LF) models struggle with ambiguous date mentions in natural language utterances, leading to erroneous query execution and inaccurate results due to random selection of date-type attributes in database schemas.

Innovation Solution

A method involving augmented training data is used to train a machine learning model that encodes date ambiguities, allowing it to recognize and represent ambiguous date mentions in output logical forms, alerting downstream tasks to potential ambiguities and enabling accurate query execution.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If existing NL2LF models use random selection of date-type attributes, then the model can handle any date mention, but query accuracy deteriorates due to erroneous query execution

Engineering Contradiction:

Improveability to handle date mentionsVSAvoidquery accuracy

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent replaces the mechanical random selection process with a machine learning-based semantic analysis system. The model uses natural language processing to analyze the context of date mentions and automatically selects the appropriate date-type attribute based on semantic understanding rather than random chance, thereby maintaining versatility while improving query accuracy

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces an intermediate representation layer between the date mention and the date-type attribute selection. This intermediate layer encodes semantic information about the date mention's context and meaning, serving as a mediator that enables accurate mapping to the correct date-type attribute without direct random selection

Inventive Principle:

Principle #24Intermediary (Mediator)

2Loss of information

If the model encodes date ambiguities in output logical forms, then downstream tasks can be alerted to potential ambiguities, but the complexity of the transformation process increases

Engineering Contradiction:

Improvepreservation of date ambiguity informationVSAvoidtransformation process complexity

Core Design Contradiction:

Loss of informationVSDevice complexity

Solution Approach 1:

The patent performs preliminary encoding of date ambiguity information during the logical form transformation process itself. By embedding this information in the output logical form structure, the system prepares downstream tasks in advance without requiring separate post-processing steps, thus managing complexity while preserving information

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12573380B2Transforming natural language to a logical form

Publication Date: 2026.03.10 ORACLE INT CORP

AI summary

Techniques are disclosed herein for managing ambiguous date mentions in natural language utterances in transforming natural language utterances to logical forms by encoding the uncertainties of the ambiguous date mentions and including the encoded uncertainties in the logical forms. In a training phase, training examples including natural language utterances, logical forms, and database schema information are automatically augmented and used to train a machine learning model to convert natural language utterances to logical form. In an inference phase, input database schema information is augmented and used by the trained machine learning model to convert an input natural language utterance to logical form.