Processing Model Knowledge Domain Segmentation for Query Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine learning processing models struggle to provide accurate query results across all knowledge domains, particularly in subdivided domains, due to their inability to cover all necessary knowledge domains.
Innovation Solution
A method is proposed where a first knowledge domain within a general knowledge domain is identified, and a set of reference queries and answers are generated and obtained from a repository associated with the processing model. These reference queries and answers are then used to update the processing model, enriching its knowledge reserve in subdivided domains.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a processing model covers all knowledge domains to provide accurate query results, then the accuracy of query results improves, but the complexity of the model increases
Solution Approach 1:
The patent divides the general knowledge domain into multiple subdivided knowledge domains (e.g., mathematics, physics, chemistry, biology, geography, history, politics, economics, law, medicine, engineering, computer science, arts, sports). Instead of requiring the model to learn all knowledge domains simultaneously, the system segments the knowledge and processes each domain separately through domain-specific reference queries and answers, reducing the complexity burden while maintaining comprehensive coverage.
Solution Approach 2:
The patent introduces a new dimension of domain-specific reference data (reference queries and reference answers) as an additional layer of knowledge enhancement. This reference data dimension complements the model's existing knowledge by providing authoritative, domain-specific information that improves accuracy without requiring fundamental changes to the model's core architecture.
2Measurement precision
If domain-specific reference queries and answers are used to update the processing model, then the accuracy in subdivided domains improves, but the processing time and computational resources increase
Solution Approach 1:
The system performs preliminary actions by pre-generating domain-specific reference queries and collecting corresponding reference answers from authoritative sources before model updating. This preparation work is done in advance, organizing knowledge into structured reference data that can be efficiently applied during model updates, reducing the time required for actual model training and adaptation.
Solution Approach 2:
The patent creates copies of knowledge in the form of reference queries and reference answers that mirror real-world domain knowledge. Instead of requiring the model to learn from raw, unstructured data, the system uses these copied, structured reference pairs as training data, which accelerates the learning process while maintaining high accuracy in subdivided domains.
Data Source
AI summary
Methods, apparatuses, devices, and media for model processing are provided. In an approach, a first knowledge domain in the field of general knowledge involved in a processing model is determined. A first set of reference queries associated with the first knowledge domain is generated. A first set of reference answers matching the first set of reference queries is obtained from a repository that is associated with the processing model and at least relates to the first knowledge domain. The processing model is updated by using the first set of reference queries and the first set of reference answers. With the implementations of the disclosure, the knowledge reserve in various subdivided knowledge domains can be continuously added to the processing model. The processing model can improve performance and accuracy in this way, providing more accurate query result output.