Voice Query Entity Recognition for Corrective Content Discovery

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Traditional voice search systems fail to correctly interpret user queries when command words are unrecognized or misrecognized, requiring users to repeat queries, leading to inefficiencies and user dissatisfaction.

Innovation Solution

Implement a system that recognizes a corrective query by identifying trigger terms like 'please', 'no', or 'I meant', adjusts the entity recognition model's relaxation rate, and considers phonetically similar entities, using context from previous queries to refine search results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional voice search systems use strict entity recognition to ensure search accuracy, then search precision is improved, but user satisfaction deteriorates when misrecognition occurs requiring query repetition

Engineering Contradiction:

Improvesearch accuracyVSAvoiduser satisfaction

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The entity recognition model dynamically adjusts its relaxation rate based on whether a corrective query is detected. When trigger terms like 'please' or 'no' are identified, the system temporarily increases the relaxation rate to consider phonetically similar entities, allowing flexible interpretation of potentially misrecognized terms without permanently compromising search precision

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the relaxation rate parameter of the entity recognition model conditionally. By detecting trigger terms in corrective queries, the system adjusts this parameter to expand the set of candidate entities considered, enabling recovery from misrecognition while maintaining strict recognition standards for initial queries

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If the entity recognition model considers more candidate entities to handle phonetic similarities, then content discovery is improved, but search precision deteriorates due to increased false positives

Engineering Contradiction:

Improvecontent discoveryVSAvoidsearch precision

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system dynamically controls the relaxation rate parameter based on query context. For corrective queries containing trigger terms, the relaxation rate is temporarily increased to improve content discovery by considering phonetically similar entities. For regular queries, the standard stricter recognition threshold is maintained to preserve search precision

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system uses feedback from trigger term detection to adjust entity recognition behavior. When users provide corrective feedback through trigger terms like 'please' or 'no', the system interprets this as permission to relax recognition standards and consider alternative phonetic matches, using the user's own feedback to guide the precision-versus-recall tradeoff

Inventive Principle:

Principle #23Feedback

3Measurement precision

If the system requires users to repeat queries with correct command terms, then search accuracy is maintained, but time efficiency deteriorates due to additional query iterations

Engineering Contradiction:

Improvecommand recognition accuracyVSAvoidquery correction time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary analysis of corrective queries by detecting trigger terms before full entity recognition. This preliminary action identifies opportunities for phonetic matching early in the processing pipeline, allowing the system to correct misrecognized commands within the same query interaction rather than requiring separate correction queries

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The relaxation rate parameter acts as an intermediary mechanism between strict entity recognition and phonetic matching. By adjusting this parameter based on trigger term detection, the system mediates between maintaining accuracy and accepting phonetically similar alternatives, enabling command correction without requiring users to explicitly repeat queries with correct terminology

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20250239255A1Systems and methods for improving content discovery in response to a voice query using a recognition rate which depends on detected trigger terms

Publication Date: 2025.07.24 ADEIA GUIDES INC

AI summary

A transcription of a query for content discovery is generated, and a context of the query is identified, as well as a first plurality of candidate entities to which the query refers. A search is performed based on the context of the query and the first plurality of candidate entities, and results are generated for output. A transcription of a second voice query is then generated, and it is determined whether the second transcription includes a trigger term indicating a corrective query. If so, the context of the first query is retrieved. A second term of the second query similar to a term of the first query is identified, and a second plurality of candidate entities to which the second term refers is determined. A second search is performed based on the second plurality of candidates and the context, and new search results are generated for output.