Search Query Suggestions Using Historical Click Relationships
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search interfaces fail to suggest queries that minimize search session length by predicting final stage queries based on historical data, often providing non-informative results that do not help users understand the subject of their initial query.
Innovation Solution
A system that analyzes historical query data to determine relationship scores and click rates, generating suggested queries that are likely to be of interest to the user and reduce search session length by predicting final stage queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If the search interface provides suggested queries based on simple keyword matching, then the device complexity is reduced, but the relevance and usefulness of suggested queries deteriorates
Solution Approach 1:
The system pre-processes historical query data offline to build relationship scores and click rate statistics before they are needed for suggestions. This preliminary action stores pre-computed relationships between keywords and queries, allowing the search interface to quickly retrieve and display relevant suggestions without performing complex real-time analysis, thus maintaining low interface complexity while ensuring high suggestion relevance
Solution Approach 2:
The patent introduces an intermediary layer of pre-computed relationship scores and click rate data that mediates between the simple keyword input and the complex relevance assessment. This intermediary layer contains pre-analyzed relationships between keywords, queries, and click behaviors, allowing the system to provide relevant suggestions through simple matching operations rather than complex real-time analysis
2Measurement precision
If the system analyzes extensive historical query data to improve suggestion accuracy, then the measurement precision of query relevance improves, but the loss of time for data processing increases
Solution Approach 1:
The system performs extensive historical data analysis, relationship score computation, and click rate calculation in advance during offline processing. This preliminary action transforms raw historical query data into pre-computed relationship metrics that can be quickly retrieved and applied during live search operations, achieving high measurement precision without real-time processing delays
Solution Approach 2:
The patent segments the data processing into distinct phases: offline pre-processing of historical data to compute relationship scores and click rates, and online retrieval of pre-computed metrics for immediate suggestion generation. This segmentation allows extensive analysis to occur during offline periods when user time is not impacted, while online operations benefit from pre-computed precision without time loss
3Adaptability or versatility
If the search interface displays multiple suggested queries, then the adaptability of the search system improves, but the ease of operation deteriorates due to information overload
Solution Approach 1:
The system applies local quality by providing different numbers and types of suggestions in different contexts within the search interface. Rather than uniformly displaying the same number of suggestions for all queries, the interface adapts the quantity and prominence of suggestions based on local conditions such as query specificity, user behavior patterns, and contextual relevance, making the interface simple to use while maintaining high adaptability
Solution Approach 2:
The patent implements partial action by selectively displaying only the most relevant suggested queries rather than all possible suggestions. The system uses pre-computed relationship scores and click rates to filter and prioritize suggestions, presenting a manageable subset that balances adaptability with ease of operation by avoiding information overload while still providing versatile search options
Data Source
AI summary
One or more computing devices, systems, and/or methods for generating a list of suggested queries associated with one or more keywords are provided. For example, one or more keywords may be received via a search interface. A plurality of queries associated with the one or more keywords may be determined based upon the one or more keywords and a historical query database. A plurality of relationship scores associated with the plurality of queries may be generated based upon a plurality of search sessions associated with the historical query database. The historical query database may be analyzed to determine a plurality of click rates associated with the plurality of queries. A list of suggested queries may be generated based upon the plurality of relationship scores and the plurality of click rates.