Chatbot Secondary Classification for Unclassified Message Handling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Chatbots typically limit their message handling to specific topics and ignore or respond with defaults to messages outside their configured characteristics, which can be undesirable in certain situations.
Innovation Solution
A system and method that utilizes a secondary classification database to enable chatbots to respond to messages not classified by primary classification data, allowing them to handle a broader range of topics and provide appropriate responses, including emergency or safety-related messages, by interconnecting with a messaging server that stores and updates secondary classification data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If chatbots are configured to recognize only specific message characteristics, then their response accuracy for configured topics is improved, but their ability to handle unexpected or critical messages deteriorates
Solution Approach 1:
The patent segments message handling into two independent classification systems: primary classification for configured topics and secondary classification for unexpected messages. This segmentation allows the chatbot to maintain precise responses for known topics while simultaneously capturing and responding to unexpected messages through the secondary system, resolving the contradiction between specialization and versatility.
Solution Approach 2:
The patent introduces a secondary classification system as an intermediary layer between the primary classification system and the message response generation. This intermediary captures messages that escape primary classification, ensuring they are not simply ignored but are instead processed through additional analysis to generate appropriate responses, thereby bridging the gap between configured topics and unexpected messages.
2Ease of operation
If chatbots ignore messages not matching configured characteristics, then their operational simplicity is maintained, but their reliability in handling critical messages deteriorates
Solution Approach 1:
The patent implements preliminary action by continuously monitoring messages through the secondary classification system before finalizing the response decision. This preliminary screening ensures that critical or unexpected messages are captured and processed in advance, preventing them from being simply ignored while maintaining the overall simplicity of the chatbot's operational framework.
Solution Approach 2:
The patent incorporates feedback mechanisms where the secondary classification system provides feedback about unexpected or critical messages to the primary response generation system. This feedback loop ensures that messages which would otherwise be ignored are fed back into the response generation process, improving reliability without significantly complicating the chatbot's operation.
3Ease of operation
If chatbots use default responses for unclassified messages, then their operational simplicity is preserved, but the quality of response to unexpected messages deteriorates
Solution Approach 1:
The patent introduces dynamics by making the response generation process adaptive rather than static. The secondary classification system dynamically adjusts the response generation process for unexpected messages, applying enhanced analysis and more precise response formulation only when needed, rather than applying full precision to all messages uniformly. This maintains simplicity for routine messages while improving precision for critical ones.
Solution Approach 2:
The patent applies parameter changes by modifying the response generation parameters based on the classification results. For messages classified by the secondary system as unexpected or critical, the system changes parameters such as response depth, analysis intensity, and formulation precision. This selective parameter adjustment improves response quality for unexpected messages without significantly increasing the computational overhead for routine messages.
Data Source
AI summary
A computing device executing a chatbot application stores primary classification data in a memory; the primary classification data includes records each containing a primary class attribute and corresponding primary response data. The device obtains and stores, from a central repository, a copy of secondary classification data including records each containing a secondary class attribute and corresponding secondary response data. The device receives a message from a client device, and determines whether the message matches any of the primary class attributes. When the message does not match any of the primary class attributes, the device determines whether the message matches any of the secondary class attributes. Based on a match between the message and one of the secondary class attributes, the device selects secondary response data corresponding to the one of the secondary class attributes; and transmits a response to the client device, including the selected secondary response data.