Voice Assistant Query Processing with Intent and Retrieval Modules
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current voice assistant devices struggle to properly interpret and respond to e-commerce related questions, especially those without entities, with complex or varied syntax, and those requiring knowledge lookup, as they often fail to understand the intent and require external information.
Innovation Solution
A computing device determines responses to user queries by identifying key and corresponding key values, calculating similarity values, and generating responses based on matching keys, utilizing intent models, word vectors, and featurization vectors to parse and respond to verbal queries effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If current voice assistant devices use basic intent identification to process queries, then simple transactional queries can be handled, but queries requiring external information or knowledge lookup cannot be properly responded to
Solution Approach 1:
The patent introduces an intermediary processing layer that includes a query classification module and an information retrieval module. This intermediary layer determines whether a query requires external information lookup before generating a response, acting as a mediator between the basic intent identification and the final response generation, thereby improving both versatility and reliability
Solution Approach 2:
The patent segments the query processing into distinct modules: intent identification, query classification, information retrieval determination, and response generation. By dividing the processing pipeline into separate functional segments, the system can selectively apply different processing paths based on query requirements, enhancing adaptability without compromising response accuracy
2Ease of operation
If voice assistant devices attempt to interpret all types of queries using the same method, then consistent processing can be maintained, but queries with complex or varied syntax fail to be properly understood
Solution Approach 1:
The patent implements dynamic query processing where the system adapts its interpretation method based on the query characteristics. The query classification module dynamically determines the appropriate processing path for each query, allowing the system to maintain consistency in its decision-making framework while being adaptable to varied syntax patterns and complex query structures
3Speed
If voice assistant devices use simple intent identification, then processing speed can be maintained, but queries without entities or requiring knowledge lookup cannot be properly interpreted
Solution Approach 1:
The patent applies preliminary action by performing query classification and determining information retrieval needs before full query processing. This preliminary assessment allows the system to quickly route simple queries for fast processing while directing complex queries requiring external information to appropriate handling paths, maintaining speed for simple cases while improving completeness for complex cases
Data Source
AI summary
An e-commerce voice response determination system is provided and generally includes a server and a customer computing device. The server may determine, based on a user query, a plurality of product descriptions. The server may determine a similarity value between the user query and each of the product descriptions based on the similarity between each word of the user query and each word of the product descriptions. The server may determine the product description with the most similarity to the user query, and provide a response to the customer computing device based on that product description. In some embodiments, the server may determine whether the user query matches a predefined query. If the user query matches the predefined query, the server provides a response to the customer computing device based on a predefined response to the predefined query.