Dialogue Skill Classification Using Semantic Slot Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining a skill field in dialogue text often result in incorrect classification due to the lack of correlation between semantic slots, leading to poor user experience, especially when multiple skills have similar parsing confidence.
Innovation Solution
Establish a knowledge base that correlates name and character semantic slots across different skill fields, allowing for accurate matching and determination of the correct skill field by checking multiple knowledge bases if initial matching fails.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a regularization-based matching method is used to determine skill field, then the matching process is simple and fast, but the classification accuracy deteriorates when multiple skills have similar parsing confidence
Solution Approach 1:
The patent replaces the traditional regularization-based mechanical matching system with a knowledge graph-based semantic analysis system. The knowledge graph stores semantic relationships between entities (e.g., singer-song relationships) and uses semantic similarity calculation to determine skill field, substituting the simple string-matching mechanism with a more intelligent semantic understanding mechanism that resolves ambiguity when multiple skills have similar parsing confidence
Solution Approach 2:
The patent introduces a knowledge graph as an intermediary between the dialogue text and skill field determination. The knowledge graph acts as a mediator that provides semantic context and relationships (e.g., linking singer names to their songs) to disambiguate cases where multiple skills could match, thereby improving classification accuracy without significantly impacting matching speed
2Productivity
If default preferential skill is used when parsing confidence is equal, then the system can quickly resolve ambiguity, but wrong skill field is selected due to lack of semantic slot correlation
Solution Approach 1:
The patent replaces the default preferential skill mechanism (which arbitrarily selects based on predefined priority) with a knowledge graph-based semantic verification mechanism. Instead of mechanically selecting the first matching skill, the system uses semantic relationships in the knowledge graph to verify which skill field actually matches the user intent, thereby improving reliability while maintaining reasonable response speed
Solution Approach 2:
The patent implements a feedback mechanism where the knowledge graph provides semantic validation feedback to the skill field selection process. When multiple skills have similar parsing confidence, the system queries the knowledge graph for semantic relationships (e.g., does the singer actually have this song?) and uses this feedback to select the correct skill field, preventing wrong selections that would occur with simple preferential matching
Data Source
AI summary
Disclosed is a method for determining a skill field of a dialogue text including determining a skill field hit by a dialogue text input by a user, and a name semantic slot and a character semantic slot in the skill field; when the dialogue text hits a first skill field, determining whether the name semantic slot and the character semantic slot match according to a knowledge base of the first skill field; determining, if not matched, whether the name semantic slot and the character semantic slot match according to a knowledge base of a second skill field; and determining, if matched, the second skill field as the skill field of the dialogue text. Also provided is an apparatus for determining a skill field of a dialogue text. The error rate of field classification is reduced, and the skill field can be hit by the user's voice dialogue more accurately.