System

A system comprising:

• circuitry configured to:
• receive an incoming communication signal via a communication interface coupled to a packet-switched network, extract identification information from signaling data, and perform a lookup against a caller profile table stored in a storage device to determine whether the identification information is known or unknown;
• select a prompt template from a plurality of prompt templates stored in the storage device based on signaling metadata, generate a prompt sentence by substituting the identification information into a placeholder of the selected prompt template, and input the prompt sentence to a transformer-based generative neural network model to obtain a safety evaluation result comprising a safety classification token and explanation tokens;
• parse the safety classification token and compare a confidence score to a threshold parameter to generate a determination, record the determination in the storage device, and generate a connection control signal transmitted to the communication apparatus to control a connection state;
• transmit notification information comprising the identification information, the safety evaluation result, and response options to a terminal device via the communication interface, receive a user response operation, and control establishment or termination of the incoming communication based on the user response operation;
• and
• update the threshold parameter and the selected prompt template based on a running accuracy metric computed from accumulated determination results and user response operations.

A method comprising:

• receiving, via a communication interface coupled to a packet-switched network, an incoming communication signal from a communication apparatus and extracting identification information from signaling data of the incoming communication signal;
• performing a lookup of the identification information in a caller profile table stored in a storage device, and generating a prompt sentence incorporating the identification information and inputting the prompt sentence to a generative neural network model to obtain a safety evaluation result;
• analyzing the safety evaluation result to generate a determination,
• generating a connection control signal, and transmitting the connection control signal to the communication apparatus to control a connection state;
• transmitting notification information comprising the identification information, the safety evaluation result, and response options to a terminal device and controlling establishment or termination of the incoming communication based on a user response operation;
• and
• updating at least one of a prompt template and a threshold parameter based on the safety evaluation result, the determination, and the user response operation.