Semantic Content Search for Diverse Intent Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search methods are inadequate for handling diverse user search intents, often failing to provide accurate and diverse search results due to their reliance on clear objectives and similarity-based rankings.
Innovation Solution
A content search method involving artificial intelligence that extracts text and content features, maps them to semantic relevance, performs semantic recognition, and groups features by type to determine relevant search results, using neural networks and contrastive learning to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional similarity-based search methods are used, then search speed and simplicity are maintained, but search result accuracy and diversity deteriorate when user search intents are diverse
Solution Approach 1:
The patent segments the search process into multiple stages: initial similarity-based filtering followed by semantic type recognition and re-ranking. This segmentation allows the system to handle diverse search intents by applying different evaluation criteria at different stages, improving both accuracy and adaptability simultaneously
Solution Approach 2:
The patent introduces dynamic adaptability by enabling the search system to adjust its evaluation methodology based on detected semantic types. When diverse search intents are identified, the system dynamically switches from pure similarity ranking to semantic-type-aware ranking, resolving the contradiction between maintaining simplicity and handling diversity
2Measurement precision
If semantic recognition and grouping operations are added to improve search accuracy, then search result quality improves, but computational complexity increases
Solution Approach 1:
The patent applies preliminary action by performing semantic type recognition on all candidate results before the final ranking stage. This preliminary classification organizes the data in a way that simplifies subsequent re-ranking operations, achieving high accuracy without proportionally increasing overall system complexity
Solution Approach 2:
The patent introduces semantic type labels as an intermediary layer between raw similarity scores and final ranking decisions. This intermediary representation simplifies the complex task of handling diverse search intents by providing a structured intermediate form that is easier to process and combine with original similarity metrics
Data Source
AI summary
A content search method performed by a computer device includes: obtaining search information and a media resource including a plurality of pieces of media content; extracting a text feature from the search information and a content feature from each of the plurality pieces of media content; transforming the plurality of content features to multiple mapped features, wherein a distance between a pair of mapped features represents semantic relevance between the pair of mapped features; performing semantic recognition on the mapped features based on the text feature, to determine semantic types corresponding to the mapped features; grouping the mapped features corresponding to the same semantic type into a same combination, and determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations; and determining search results for the search information from the media resource according to the target mapped features.