APPARATUS AND METHOD FOR DETECTING UNETHICAL LANGUAGE BASED ON LEARNING

摘要

In the present invention, by dividing sentences into character units, it is possible to detect unethical words robustly even in grammatical errors such as spacing or spelling, and when character-specific features are extracted, previous features are cumulatively reflected, so that they are divided into character units. Accordingly, it is possible to provide an unethical detection apparatus and method capable of accurately determining whether an unethical is included even though it is difficult to analyze the meaning of a sentence.