Query Expansion Using Add-On Terms With Classifications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current query expansion methods in information retrieval, such as web and advertising searching, face challenges in disambiguating between different contexts for query terms and accommodating directly related terms, leading to computationally expensive multiple queries and increased server load.
Innovation Solution
The method involves identifying and appending add-on terms with assigned classifications to the original query, allowing for a single reformulated query to be executed on a resource stack, where each add-on term retains its classification, distinguishing relevant resources and reducing server load by eliminating the need for multiple queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple queries are executed to expand query terms, then retrieval performance improves, but server load and computational expense increase
Solution Approach 1:
The patent merges multiple query expansions into a single query by appending add-on terms with classifications to the original query. Instead of executing multiple separate queries for different query terms, the system combines them into one unified query structure that preserves context distinctions through classifications, thereby reducing server load while maintaining retrieval performance.
Solution Approach 2:
The patent segments the query into the original query term and separate add-on terms, each with assigned classifications. This segmentation allows the system to distinguish between different contexts and meanings of query terms while executing a single query, preventing the need for multiple separate queries and reducing computational expense.
2Reliability
If multiple queries are executed to expand query terms, then retrieval performance improves, but computational expense increases
Solution Approach 1:
The patent merges multiple query expansions into a single query by appending add-on terms with classifications to the original query. Instead of executing multiple separate queries for different query terms, the system combines them into one unified query structure that preserves context distinctions through classifications, thereby reducing server load while maintaining retrieval performance.
3Reliability
If query terms are expanded without classifications, then retrieval performance improves, but context distinction is lost
Solution Approach 1:
The patent applies local quality by assigning specific classifications to individual add-on terms within the query. Each add-on term receives a classification that captures its specific context and relationship to the original query term. This allows the system to maintain precise context distinction for each term while still executing a single unified query, preventing information loss.
Data Source
AI summary
In various embodiments, systems and methods are provided for query expansion using add-on terms with classifications. A query is received. An add-on term is identified for the query. A classification is determined for the add-on term. The classification is a designation associated with the add-on term that is used to distinguish the add-on term from the query. An appended query is generated based on the add-on term. The appended query is generated by concatenating the query with the add-on term. The appended query is executed on a resource stack as a single reformulated query to identify one or more resources. Upon execution, the classification of the add-on term distinguishes the one or more resources identified for the add-on term based on tagging the one or more resources with the classification of the add-on term. The appended query is used to generate content items.