Crowdsourced Parsing Rule Generation for Speech Command Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Developing parsing rules for speech command processing in speech recognition systems is resource-intensive due to the numerous possible parses of input sentences, making it challenging to efficiently recognize and perform user actions.
Innovation Solution
A system utilizing crowdsourcing techniques to generate parsing rules through user interfaces that allow users to input and classify command sentences, enabling the creation of command grammars that facilitate a bottom-up parse of input sentences to determine invoked actions, without requiring users to have a detailed understanding of grammars.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional methods are used to develop parsing rules for speech commands, then action recognition can be achieved, but the process is resource-intensive and time-consuming
Solution Approach 1:
The system enables users to self-service by providing intuitive interfaces where users can input command sentences and classify n-grams without requiring detailed grammar knowledge. The system automatically learns parsing rules from these user inputs, eliminating the need for expert linguists to manually create parsing rules, thus dramatically improving productivity and reducing time loss.
Solution Approach 2:
The patent replaces the mechanical process of manual parsing rule creation by experts with an automated machine learning system. The system uses computational algorithms to process user inputs, classify n-grams, and generate parsing rules automatically, substituting the manual mechanical process with an automated digital system that is both faster and more scalable.
2Measurement precision
If detailed grammar knowledge is required for parsing rule development, then parsing accuracy can be maintained, but user accessibility and ease of contribution are reduced
Solution Approach 1:
The system segments the complex task of parsing rule development into simpler, more manageable components. Instead of requiring users to understand entire grammars, the system breaks down sentences into n-grams and classifies each segment independently. This segmentation makes the contribution process accessible to ordinary users while maintaining overall parsing accuracy through the aggregation of these simplified classifications.
Solution Approach 2:
The patent introduces an intermediary layer between the user and the complex grammar system. The n-gram classification interface serves as a mediator that translates simple user inputs into structured parsing rules. This intermediary abstracts away the complexity of formal grammars, allowing users to contribute without detailed knowledge while the system maintains parsing precision through the intermediary's processing.
3Adaptability or versatility
If multiple possible parses are considered for each input sentence, then action recognition completeness is improved, but system complexity and resource requirements increase
Solution Approach 1:
The system applies partial action by focusing on classifying only the most relevant n-grams rather than exhaustively analyzing every possible parse. By identifying and classifying key n-gram segments, the system achieves sufficient action recognition completeness without the computational burden of exhaustively processing all possible parses, thus reducing system complexity while maintaining versatility.
Data Source
AI summary
Systems, methods and apparatus for learning parsing rules and argument identification from crowdsourcing of proposed command inputs. Crowdsourcing techniques to generate rules for parsing input sentences. A parse is used to determine whether the input sentence invokes a specific action, and if so, what arguments are to be passed to the invocation of the action.