Vector Aggregation for Search Intent Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Search engines face challenges in accurately identifying user intent behind search queries, leading to suboptimal search result retrieval and navigation due to variations in query intent expression and the large number of possible search results, which conventional methods fail to address effectively.
Innovation Solution
The technology employs interaction tracking and aggregation of item description vectors to better understand user intent by mapping search queries to relevant search results, using methods like averaging and neural networks to identify related search queries that reflect user intent more accurately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional search methods are used to identify user intent, then the search process is simple and fast, but the accuracy of identifying user intent deteriorates leading to suboptimal search results
Solution Approach 1:
The system pre-computes and stores aggregate item description vectors for multiple search queries before runtime. This preliminary action creates a ready-to-use reference database that enables fast and accurate intent identification during actual search operations without requiring complex real-time computations.
Solution Approach 2:
The patent introduces aggregate item description vectors as an intermediary representation between search queries and search results. These vectors serve as a mediator that captures the semantic meaning of queries and enables comparison with item descriptions, improving intent identification accuracy without directly increasing system complexity.
2Adaptability or versatility
If the number of search results is increased to provide more options, then user navigation capability improves, but computational intensity increases making the system slower
Solution Approach 1:
The system performs preliminary aggregation of item description vectors during an offline phase, creating a compact representation of search results. This pre-computation reduces the computational burden during runtime, allowing the system to handle large numbers of search results efficiently without excessive computational intensity.
Solution Approach 2:
The patent extracts the essential semantic information from numerous search results by aggregating their item description vectors into a single representative vector. This extraction process captures the core meaning while discarding redundant details, enabling the system to work with condensed representations that reduce computational requirements.
3Speed
If aggregate item description vectors are pre-computed and stored, then search query matching speed improves, but memory requirements increase
Solution Approach 1:
The system extracts only the essential aggregate item description vectors that represent the semantic meaning of search results. By storing only these condensed vector representations rather than full search result data, the system achieves fast matching speed while minimizing memory storage requirements.
Solution Approach 2:
The patent transforms detailed search result information into vector representations with a fixed, optimized dimensionality. This parameter transformation converts variable-sized text data into standardized vector formats, enabling efficient storage and fast comparison operations while controlling memory usage.
Data Source
AI summary
Search queries are received and search results are provided. Interaction tracking is used to determine with which search results users interact. The search results having received interactions can be represented as item vectors, which can include a vector representation of a portion of the search result, such as a title, description, or image. For each search query, the item vectors are aggregated, such as by averaging the item vectors. The search queries are stored in an item dataset as collected search queries respectively associated with the aggregate item vectors. When a new search query is received, a search query vector can be compared to the aggregate item description vectors to identify collected search queries that are related. The related collected search queries can be provided as search query recommendations or search results associated with the collected search queries can be provided in response to receiving the new search query.