Semantic Fusion of Query Logs and Ontologies for Object Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic information classification methods face challenges in accuracy and user-friendliness due to their reliance on either query histories or ontological information alone, with query log-based methods lacking background knowledge and ontology-based methods being inflexible and unable to reflect changes in user interests.
Innovation Solution
A method and system that combine query log-based and ontological information-based classification results through semantic fusion, involving three steps: query log-based classification, ontology-based classification, and the semantic combination of the two results to generate a final classification result, allowing for improved accuracy and user-friendly display.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If query log-based classification is used, then the classification reflects user interests, but the classification accuracy is insufficient due to lack of background knowledge
Solution Approach 1:
The patent combines query log-based classification results with ontological information-based classification results through semantic fusion. The query log component captures user interests and behavioral patterns, while the ontology component provides structured background knowledge and semantic relationships. By merging these two classification approaches, the system achieves both adaptability to user interests and accuracy from ontological knowledge.
2Measurement precision
If ontological information-based classification is used, then the classification accuracy improves through background knowledge, but the system becomes inflexible and cannot reflect changes in user interests
Solution Approach 1:
The patent introduces dynamic elements into the ontology-based classification by incorporating query log data that reflects real-time user interests and search behaviors. While the ontological structure remains static and provides stable background knowledge, the query log component dynamically adapts to changing user preferences. The semantic fusion mechanism dynamically weights and combines these two sources, allowing the system to be both accurate and adaptable.
3Ease of operation
If query log-based classification is used, then the classification is user-friendly, but the classification accuracy is not good enough without background knowledge
Solution Approach 1:
The patent merges query log-based classification (which provides user-friendliness by reflecting actual user search patterns and interests) with ontological information-based classification (which provides accuracy through structured background knowledge). The semantic fusion process integrates these two approaches, delivering classification results that are both accurate and user-friendly.
4Measurement precision
If ontological information-based classification is used, then the classification is accurate, but the category set is inflexible and cannot reflect changes in user interests
Solution Approach 1:
The patent makes the category set dynamic by combining the static ontological categories with dynamic query log-derived categories. The ontological information provides a stable foundation of accurate categories, while the query log component introduces flexibility by capturing emerging user interests and trending topics. The semantic fusion mechanism dynamically adjusts the category set based on current user behavior patterns.
Data Source
AI summary
The present invention provides a method and system for automatic objects classification. The method comprises: acquiring a set of objects; classifying the objects based on query log to generate a first classification result; classifying the objects based on ontological information to generate a second classification result; and semantically fusing the first and second classification results to generate a final classification result. According to the present invention, compared with the prior arts, by semantically fusing the query log-based classification result and the ontology-based classification result, the accuracy and user-friendness of the object classification can be improved.