abstract |
The present application discloses a method and device for realizing security auditing of text content. By acquiring auditing content, it is determined whether the auditing content has an auditing record according to a content ID; if the page of the auditing content is normal and there is no auditing record, the Perform keyword filtering on the audit content; if there are no sensitive words in the audit content after keyword filtering, determine whether the audit content has department-sensitive words; if there are no department-sensitive words, split the audit content and obtain the dismantled content. Post-divided content; AI review of the divided content to obtain review results, quick review of the content of user questions and doctor’s answers in the medical and health field, and prioritizing the modification of any illegal or indecent content, If it cannot be modified, the deletion operation is adopted, so that the text content not only meets the audit requirements, purifies the website content, but also reduces the display of colloquial and indecent content, making the content more in line with the reading habits of users. |