Search Result Sorting Model Using Query Vector Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional search engines face inaccuracies in entity recommendations due to ambiguity in user queries, as they only consider the current query without accounting for historical context, leading to unclear recommendations for terms like 'Apple' which could refer to a fruit, movie, or corporation.
Innovation Solution
A method and apparatus that utilize a search result-sorting model to integrate current and historical query vectors, employing neural networks and attention mechanisms to score candidate search results based on similarity, thereby improving the relevance of search results by considering the historical query sequence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional search engines only consider the current query for entity recommendations, then the system complexity remains low, but the accuracy of search result recommendations deteriorates due to query ambiguity
Solution Approach 1:
The system pre-processes historical queries and transforms them into vector representations before the actual search occurs. By preparing the historical context in advance and integrating it with the current query vectors, the system resolves query ambiguities before presenting search results, thereby improving recommendation accuracy without adding significant complexity during the actual search operation
Solution Approach 2:
The patent introduces vector representations as an intermediary between the query text and the search results. By transforming both historical queries and current queries into vector space, the system can compute similarity measures and integrate contextual information without directly processing complex textual relationships, thus improving accuracy while maintaining manageable system complexity
2Measurement precision
If the system integrates historical query sequences to resolve query ambiguity, then the accuracy of entity recommendations improves, but the computational requirements and processing time increase
Solution Approach 1:
The system pre-computes and stores vector representations of historical queries in advance. When a current query arrives, the system quickly retrieves and integrates these pre-processed vectors rather than re-processing historical data, significantly reducing the time penalty associated with considering historical context while maintaining improved recommendation accuracy
Solution Approach 2:
The system computes vector representations and similarity measures for only the most relevant historical queries rather than processing the entire historical query sequence. By selectively integrating partial historical context that is most useful for resolving the current query's ambiguity, the system achieves improved accuracy without the full computational cost of analyzing all historical data
Data Source
AI summary
The present disclosure provides a method and apparatus for determining search results, a device and a computer storage medium, wherein the method comprises: obtaining a current query, a historical query sequence of the current query and candidate search results of the current query; inputting the current query, the historical query sequence of the current query and candidate search results of the current query into a search result-sorting model, scoring the candidate search results according to the search result-sorting model, and determining search results corresponding to the current query; wherein the scores rendered by the search result-sorting model for the candidate search results are determined according to a similarity degree between an integrated vector representation of the current query and the historical query sequence of the current query and vector representations of the candidate search results. The present disclosure can provide the user with search results which more accurately reflect the user's demands.