Frequently Asked Questions (FAQs)

The interaction between AI-created exams and the editing page is as follows: 1. After a successful import, it automatically redirects to the questionnaire editing page 2. The imported score settings will be displayed correctly on the editing page 3. For true/false questions, if no options are set, two options, "True/False," will be added by default 4. The editing page will display the correct answers set during AI generation 5. From the Excel page, you can quickly access the AI creation feature through the "AI Create" entry 6. When copying a questionnaire, the AI-created exam entry will be retained

Characteristics of content handling when AI creates exams in different language environments: 1. The system language and the AI creation language can be different; for example, a Chinese exam can be created in a Japanese system environment. 2. When the language of the uploaded document differs from the selected questionnaire language, the system will attempt to translate the content into the target language. 3. The system automatically handles question type names and answer formats across different language environments. 4. In some language combinations (for example, Japanese system language and Korean questionnaire creation), the generated content may display a mix of both languages. 5. When the document content does not match the selected questionnaire language, the system will try its best to generate questions according to the selected language.

Performance and limitations of the AI exam creation feature: 1. The system can stably handle the generation of about 50 questions 2. The text input limit is 30,000 characters (increased from the previous 5,000 characters) 3. The file size limit is 2MB (adjusted from the previous 5MB) 4. The generation process usually takes some time, and the system will show progress prompts 5. The number of generated question types may not exactly match the settings, but the system will try to keep it within the specified range 6. Some complex documents may not be able to extract valid text information