System

A system comprising:

• circuitry configured to:
• acquire character string data representing a posting draft from a user terminal;
• preprocess the character string data using a natural language processing component to segment the character string data into linguistic units;
• compare the preprocessed character string data with dictionary information and regular expression information to determine presence or absence of offensive expressions or confidential information and generate detection result data comprising a determination result and a risk level;
• construct a prompt sentence based on the detection result data and transmit the prompt sentence to a generative artificial intelligence model to obtain response character string data comprising a candidate modification proposal;
• re-compare the response character string data with the dictionary information and the regular expression information, generate display control information specifying problem portions as display position information, and transmit presentation information comprising a warning message and the candidate modification proposal to the user terminal;
• and
• apply an emotion recognition model to at least one of the character string data and user input information to determine an emotional state and adjust at least one of the warning message and the candidate modification proposal based on the emotional state.

A method performed by circuitry, the method comprising:

• acquiring character string data representing a posting draft from a user terminal;
• preprocessing the character string data using a natural language processing component to segment the character string data into morpheme units or phrase units;
• comparing the preprocessed character string data with dictionary information and regular expression information to determine presence or absence of offensive expressions or confidential information and generating detection result data comprising a determination result and a risk level;
• constructing a prompt sentence defining warning content and a modification policy based on the detection result data and transmitting the prompt sentence to a generative artificial intelligence model to acquire response character string data comprising a candidate modification proposal;
• re-comparing the response character string data with the dictionary information and the regular expression information;
• applying an emotion recognition model to at least one of the character string data and user input information to determine an emotional state and adjusting at least one of the warning message and the candidate modification proposal based on the emotional state;
• and
• generating posting permission information or additional warning information based on whether modified character string data satisfies a predetermined safety criterion.