Search Query Phrase Segmentation and Annotation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional search engines face difficulties in determining relationships between terms in search queries, which affects the quality of search results presented to users, as they often treat individual terms independently rather than as phrases.
Innovation Solution
A method is introduced to segment search queries into phrases and annotate them, using a predictive model trained with authoritative and behavioral data to determine the best segmentation, allowing search engines to interpret terms as phrases for more accurate resource retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If search engines treat individual terms independently in search queries, then the processing simplicity is maintained, but the search result quality deteriorates due to inability to determine relationships between terms
Solution Approach 1:
The patent segments the search query into multiple phrases by identifying term boundaries and relationships. The query processing system divides the query into phrases based on semantic relationships between terms, allowing each phrase to be processed and ranked independently while maintaining the contextual relationships between phrases. This segmentation enables the system to capture term relationships without requiring complete analysis of all possible term combinations.
Solution Approach 2:
The patent applies preliminary action by pre-processing the search query to identify and segment phrases before the main search ranking process. The system analyzes term relationships, identifies phrase boundaries, and prepares phrase-level representations in advance. This preliminary segmentation allows the search engine to efficiently process queries without requiring complex real-time analysis of all term relationships during the ranking phase.
2Measurement precision
If search engines segment queries into phrases to improve relationship understanding, then search result accuracy improves, but the computational resources and processing time increase
Solution Approach 1:
The patent applies partial action by segmenting queries into phrases rather than analyzing every possible term combination or treating the entire query as a single unit. The system identifies and processes only the necessary phrase-level relationships, avoiding unnecessary computation on term combinations that do not form coherent phrases. This partial segmentation approach provides sufficient accuracy for search results while reducing computational overhead compared to exhaustive analysis.
3Productivity
If search engines treat all terms as independent units, then the processing speed is maintained, but the ability to identify relevant resources deteriorates
Solution Approach 1:
The patent segments the query into phrases that preserve semantic relationships between terms. By organizing terms into meaningful phrases rather than treating them as independent units, the search engine can maintain processing efficiency while improving resource relevance. The segmented phrase structure allows for optimized indexing and retrieval operations that leverage the semantic relationships between terms within each phrase.
Solution Approach 2:
The patent applies local quality by assigning different processing characteristics to different parts of the query. Each phrase is processed and ranked with consideration of its internal term relationships, while the overall query structure maintains efficiency through phrase-level organization. This local analysis of phrase relationships improves resource relevance without requiring complete reprocessing of the entire query at higher computational cost.
Data Source
AI summary
A computing device can generate a collection of phrases using both authoritative data and behavioral data, for example, using previously submitted search queries. The collection of phrases can be used, in part, to determine the best segmentation of search queries. Each segmentation of a search query splits the terms in search query using different permutations or n-grams to identify one or more phrases. Each segmentation is scored based on various criteria. The segmentation having the highest score is included in training data for training a predictive model that predicts segmentations for new search queries. The predicted segmentation can be used to annotate that query to identify the one or more phrases that were created by the segmentation of the query. The annotated query can be processed, for example, by a search engine, to obtain resources that are responsive to the one or more phrases that were identified by the segmentation.