Inference Device Knowledge Graph Update
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional inference devices fail to provide desirable results for users as they do not adequately update their knowledge base based on user responses, leading to repetitive and unsatisfactory interactions in dialogic scenarios.
Innovation Solution
An inference device that acquires situation and response information, updates its knowledge base by assigning user requests as start nodes in a knowledge graph, calculates page rank values, and adjusts edge weights to ensure the desired inference result matches user preferences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional inference techniques are used with fixed knowledge base, then inference processing is simpler, but the inference results do not match user preferences and lack adaptability
Solution Approach 1:
The patent implements feedback by acquiring user responses to inference results and using this feedback to update the knowledge base. The processor receives user responses, determines whether the inference result matched user intent, and updates edge weights in the knowledge graph accordingly. This creates a closed-loop system where inference outcomes feed back into improving future inferences, enabling the system to adapt to user preferences over time.
Solution Approach 2:
The patent applies preliminary action by pre-calculating and storing page rank values for nodes in the knowledge graph. Instead of performing complex inference calculations in real-time, the system pre-computes importance metrics and stores them in the knowledge base. This allows faster inference processing while maintaining accuracy, as the pre-computed values are readily available when needed for making inferences.
2Measurement precision
If knowledge base is updated based on user responses, then inference accuracy improves, but processing time increases due to additional update operations
Solution Approach 1:
The patent applies partial action by selectively updating only the specific edge weights in the knowledge graph that are relevant to the inference that just occurred. Instead of re-computing or re-processing the entire knowledge base, the system updates only the affected edges based on user feedback. This minimizes the time and computational resources required for updates while still improving inference accuracy for future queries.
3Measurement precision
If page rank calculation is used to determine node importance, then inference result relevance improves, but computational complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-calculating page rank values for all nodes in the knowledge graph and storing these values in the knowledge base. The pre-computation of page rank metrics eliminates the need for complex real-time calculations during inference. The stored page rank values are then directly used to determine node importance and select inference results, significantly reducing computational complexity during operation while maintaining measurement accuracy.
Data Source
AI summary
An inference device includes a knowledge base unit that stores a knowledge base for inferring answers to requests; an information acquisition unit that acquires situation information that is information including a user request or a request from a user and response information including a user's response to an inference result or a result of inference; an inference unit that uses the knowledge base to make inferences for a user request included in the situation information; an information output unit that outputs an inference result; a feedback information extraction unit that specifies a desired result or a result desired by the user for the user request from the response information; and an update unit that updates the knowledge base such that the desired result more likely matches the inference result.