Voice Query Interpretation Using Entity Feasibility Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search engines often provide irrelevant search results due to a lack of understanding of search queries, leading to broad and untargeted responses.
Innovation Solution
A method and system for interpreting voice queries by determining voice recognition terms, identifying entity names, calculating feasibility scores, and ranking terms based on these scores to execute queries in a search domain, using entity information and contextual analysis to refine search results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If search engines provide broad search results based on keyword matching, then the quantity of search results increases, but the relevance and accuracy of search results deteriorates
Solution Approach 1:
The system performs preliminary entity recognition and disambiguation on search queries before executing the search. By identifying entities and their types in advance, the system can filter and refine search results to only include relevant items, thus maintaining high quantity while improving relevance through pre-processing the query understanding
Solution Approach 2:
The patent introduces an intermediary layer between keyword matching and result retrieval. This intermediary performs entity recognition, type classification, and disambiguation to bridge the gap between simple keyword search and meaningful result retrieval, thereby improving relevance without sacrificing result quantity
2Device complexity
If search engines use simple keyword matching, then the complexity of the search system is reduced, but the understanding of search queries deteriorates
Solution Approach 1:
The search system is segmented into distinct functional modules: entity recognition module, entity type classification module, disambiguation module, and search execution module. This segmentation allows each component to perform its specific function with manageable complexity while collectively achieving sophisticated query understanding that simple keyword matching cannot provide
3Speed
If voice recognition systems select recognition terms without feasibility scoring, then the processing speed increases, but the accuracy of query interpretation deteriorates
Solution Approach 1:
The system calculates feasibility scores for only the most likely voice recognition terms rather than all possible interpretations. By applying partial action to the top candidates, the system achieves high accuracy in query interpretation while maintaining processing speed, avoiding the need to evaluate every possible recognition outcome
Data Source
AI summary
Methods, systems, and media for interpreting queries are described herein. For example, an illustrative method may include: determining, based on a voice query received in a search domain, a first voice recognition term and a second voice recognition term; determining that at least a portion of the first voice recognition term corresponds to an entity name associated with the search domain; determining, based on the entity name, a feasibility score for the first voice recognition term; ranking, based on the feasibility score, the first voice recognition term over the second voice recognition term; and executing the voice query in the search domain using the first voice recognition term, the first voice recognition term being selected for use over the second voice recognition term based on the ranking. Corresponding systems, devices, and other implementations are also described.