Autonomous Message Generation via Dual-Classification Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current chatbot systems face challenges in accurately selecting appropriate responses to incoming messages, often resulting in either infeasible computations or substantial burdens, especially when multiple characteristics need to be considered.
Innovation Solution
The system employs a method where an application server stores primary and secondary class definitions, selects corresponding response data based on message content, and generates responses by combining these data elements, allowing for autonomous message generation that accounts for topics and emotional states independently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a chatbot system uses a large set of possible responses with multiple characteristics to accurately identify appropriate responses, then the accuracy of response selection is improved, but the computational burden becomes substantial or infeasible
Solution Approach 1:
The patent segments the response selection process into two independent classification stages: primary classification identifies the topic of the user message, and secondary classification identifies the emotional state. This segmentation divides the原本 single complex classification task into two simpler, independent tasks, reducing computational complexity while maintaining accuracy in selecting appropriate responses from the large response set.
2Measurement precision
If a chatbot system considers multiple characteristics (topic and emotional state) to generate context-specific responses, then the relevance of responses is improved, but the device complexity increases
Solution Approach 1:
The system segments the classification functionality into separate primary and secondary classifiers, where each classifier handles a specific characteristic (topic or emotional state). This modular segmentation reduces system complexity by making each classifier independent and specialized, while the combination of both classifiers delivers comprehensive context-specific response relevance.
Solution Approach 2:
The patent creates a universal response selection framework that handles multiple characteristics (topic and emotional state) through a standardized two-stage classification process. This multi-functional system can accommodate different types of messages and responses while maintaining a consistent structural approach, reducing overall system complexity despite handling multiple characteristics.
Data Source
AI summary
A method is provided in an application server, comprising: storing a plurality of primary class definitions in a memory, each primary class definition including a primary class identifier and a plurality of primary class attributes; storing a plurality of secondary class definitions in a memory, each secondary class definition including a secondary class identifier and a plurality of secondary class attributes; receiving a message from a client computing device via a network; based on the content of the message, selecting one of the primary class identifiers, and one of the secondary class identifiers; selecting primary response data corresponding to the selected primary class identifier; selecting secondary response data corresponding to the selected secondary class identifier; generating a response message by combining the primary response data and the secondary response data; and transmitting the response message to the client computing device.