Semantic Graph Search Recommendations for Uncertain User Input
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search technologies face challenges in accurately recommending candidate search terms due to uncertainty and inaccuracy in user input, leading to poor search results and inefficient user experiences.
Innovation Solution
A method involving semantic representation processing of current and neighbor nodes, determining a semantic aggregation representation, and recalling candidate search terms using a trained model to improve accuracy and relevance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional search term recommendation methods are used, then the system is simple to operate, but the accuracy of recommended search terms is poor due to uncertainty and inaccuracy in user input
Solution Approach 1:
The patent segments the search term recommendation process into distinct functional modules: a graph construction module that builds knowledge graphs from search data, a semantic representation module that encodes search terms and their relationships, and a recommendation module that generates suggestions. This segmentation allows each module to specialize in specific tasks, improving overall accuracy while making the complex system more manageable and maintainable
Solution Approach 2:
The patent introduces semantic representations as an intermediary layer between raw search terms and recommendation results. By transforming search terms into semantic vectors that capture meaning and relationships, the system bridges the gap between uncertain user input and accurate recommendations, enabling more precise matching without requiring direct comparison of raw terms
2Measurement precision
If semantic representation processing is performed on current and neighbor nodes, then the accuracy of search term recommendation is improved, but the computational time and processing complexity increase
Solution Approach 1:
The patent performs preliminary action by pre-computing semantic representations and building knowledge graphs during off-peak periods or in advance. Search terms, their neighbors, and semantic relationships are pre-processed and stored in optimized structures, allowing the recommendation system to quickly retrieve and combine pre-computed representations rather than performing heavy computation in real-time
Solution Approach 2:
The patent implements dynamic processing where the level of semantic analysis applied to each search term adjusts based on contextual factors such as search frequency, user behavior patterns, and system load. For common search terms with well-established semantic representations, the system uses faster, simplified processing paths, while novel or complex queries receive more intensive semantic analysis, optimizing the balance between accuracy and speed
Data Source
AI summary
A method for recommending a search term, an electronic device and a storage medium are provided. The method includes: acquiring an input target search term; determining a neighbor node of a current node and a relationship between the current node and the neighbor node, by taking the target search term as the current node; performing semantic representation processing on first text information of the current node and second text information of the neighbor node respectively, to obtain a first semantic representation of the current node and a second semantic representation of the neighbor node; obtaining a semantic aggregation representation of the target search term according to the first semantic representation, the second semantic representation and the relationship between the current node and the neighbor node; and recalling a candidate search term to be recommended from a set of candidate search terms, by taking the semantic aggregation representation as an index.