I plan to utilize OpenAI’s capabilities for HTML question extraction, answer retrieval, and JavaScript code generation. Specifically, I aim to use OpenAI to parse HTML files(DOM), identify questions within them, get relevant answers, and generate JavaScript code snippets to insert those answers into the HTML page’s Document (DOM). I’m eager to hear your thoughts, suggestions, and potential collaboration opportunities to bring this idea to life , thank you
To provide further context, I experimented with a vision model by supplying it with a full-screen screenshot. The model demonstrated strong question-answering capabilities, but I encountered complexity in generating the correct JavaScript code to implement the answers due to the diverse nature of the DOM structures across different web pages.
I’d like to emphasize that the DOM files I’m working with exhibit significant diversity in their formats and encompass a wide range of question types. This diversity poses a unique challenge, as I aim to ensure that the course work well on our platform after they are published. Adapting to these varying formats and question types is a crucial aspect of this project.