Chat Bot Resource Allocation for Unrelated Conversations

The patent segments conversations into two distinct categories: service-related conversations and non-service conversations. This segmentation is achieved through classification units that analyze conversation content and determine whether each conversation relates to the service menu. By dividing the conversation stream into separate categories, the system can allocate resources differently for each type, ensuring that back-end system resources are only increased in response to service-related conversations while still maintaining chat bot responsiveness for all conversations.

The patent applies local quality by implementing different resource allocation strategies for different types of conversations. Service-related conversations receive full resource allocation to maintain high service quality, while non-service conversations receive minimal or no back-end system resource allocation. This is achieved through separate resource allocation paths: one for service-related conversations that triggers back-end system resource increases, and another for non-service conversations that does not trigger resource increases, thereby maintaining service quality where needed while improving overall resource efficiency.