Semantic Recognition Accuracy via Dependency Parsing Confidence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing natural language processing technologies for semantic recognition, particularly in dependency syntax analysis, face accuracy issues due to the greedy algorithm's limitation of considering only the best transition action at each step, neglecting the influence of subsequent states, which affects the accuracy of dependency syntax analysis and subsequently semantic recognition.
Innovation Solution
A method that determines multiple dependency relationships and confidence levels through various preset set update operations, considering the influence of subsequent words, and selects an optimal dependency parsing result based on the highest sum of confidence levels to improve the accuracy of semantic recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a greedy algorithm is used to select the best transition action at each step, then the computational complexity is reduced and processing speed is improved, but the accuracy of dependency syntax analysis deteriorates due to neglecting the influence of subsequent states
Solution Approach 1:
The patent performs preliminary actions by generating multiple candidate dependency relationships and parsing results in advance, rather than selecting only the best action at each step. This allows the system to consider multiple possible subsequent states and select the optimal parsing result based on the highest sum of confidence levels, thereby improving accuracy while maintaining reasonable processing efficiency.
2Measurement precision
If multiple dependency relationships are considered through various preset set update operations, then the accuracy of semantic recognition is improved, but the device complexity and computational overhead increase
Solution Approach 1:
The patent applies partial action by generating multiple candidate dependency relationships and parsing results, then selecting the optimal one based on confidence level sums. This approach considers more possibilities than a greedy algorithm (excessive action) but avoids exhaustive enumeration of all possible parsing results (partial action), achieving a balance between accuracy improvement and computational complexity management.
Data Source
AI summary
A method for semantic recognition includes: in response to performing semantic analysis on information acquired by a terminal, a sentence to be processed is acquired. Word recognition is performed on the sentence to be processed, to obtain a plurality of words and part-of-speech information thereof. A target set update operation is determined with a pre-trained word processing model, according to a word to be processed in the set of words to be processed and part-of-speech information of the word to be processed. If a dependency relationship corresponding to the target set update operation is a first dependency relationship, through each of the plurality of preset set update operations, a respective dependency relationship of the word to be processed and a respective confidence level corresponding to the dependency relationship is determined, and a respective update of the set of words to be processed is performed.