Hi, I have an LMS application, currently developed in Python-Django.
It works all very well, students can view their classes and submit exams, but I have noticed that some students are sharing exam answers with other students and let's say they already have a model of the exams.
I want with the help of artificial intelligence, the exams to have different questions and in a different order for each student, what technology should I learn to develop something like this? I am a Python-Django developer but my focus is on web development, I have never touched anything from A.I.
What do you think about TensorFlow?
Please, I would appreciate all your ideas and opinions, thank you very much in advance.
I would recommend Ollama, it lets you run a local LLM, which can rewrite the questions. I would test carefully, and review anything generated, as models can hallucinate.
modelfile='''
FROM llama2
SYSTEM Rewrite any exam questions provided to ask the same thing with the same information, but in a slightly different format to thwart cheating.
'''
ollama.create(model='questionrewriter', modelfile=modelfile)
res = ollama.generate(model='llama2', prompt='''A 27-year-old woman comes to the office for counseling prior to conception. She states that a friend recently delivered a newborn with a neural tube defect and she wants to decrease her risk for having a child with this condition. She has no history of major medical illness and takes no medications. Physical examination shows no abnormalities. It is most appropriate to recommend that this patient begin supplementation with a vitamin that is a cofactor in which of the following processes?
(A) Biosynthesis of nucleotides
(B) Protein gamma glutamate carboxylation
(C) Scavenging of free radicals
(D) Transketolation
(E) Triglyceride lipolysis''')
print(res)
Thank you for your response.
I will check out Ollama to dig deeper and see how it would work for my project.
Again, thank you very much.
https://protyrefitters.co.uk/services/mobile-tyre-repair
All types of insect infestations including bedbugs, cockroaches, termites are controlled with 100% guarantee.
Website: https://www.tigerpestcontrol.com.bd/
By focusing on NLP and randomization techniques, you can achieve your goal of generating varied exams for each student. Start with the basics and gradually explore more advanced AI concepts as needed. With your Python-Django background, you'll find many of these technologies accessible and complementary to your existing skills.
Good luck with your project!
click here more information:https://www.talentelgia.com/solutions/lms-development-services
Advice for Integrating AI into Your LMS
Hi,
To address the issue of students sharing exam answers in your Python-Django LMS, consider the following steps to implement AI for dynamic exams:
Technologies to Learn:
TensorFlow: A powerful framework for building AI models to generate unique exam questions.
Natural Language Processing (NLP): Use libraries like NLTK or spaCy for generating and understanding questions.
Randomization Algorithms: Implement Python’s random module to randomize question order and selection.
API Integration: Utilize Django REST framework to connect your AI models with your LMS.
Implementation Steps:
Data Preparation: Collect a diverse dataset of questions.
Model Training: Train models using TensorFlow to generate questions.
Question Randomization: Develop algorithms to ensure varied question order for each student.
Integration: Create REST API endpoints in Django to serve randomized questions.
Testing: Validate the system to ensure question integrity and randomness.
By following these steps, you can enhance exam integrity and provide a fair assessment environment.
Best of luck with your project!
Wooflo Pro Team
All types of insect infestations including bedbugs, cockroaches, termites are controlled with 100% guarantee.
Website: https://www.tigerpestcontrol.com.bd/
Yes, use local LLMs. No need for Tensorflow (to train something custom). You will use the HuggingFace transformers library for inferencing. You're dealing with sensitive educational data so best to not send it over servers via APIs. Unless you're experienced in security, but even then your institution will probably not allow it.
You can use LM Studio to serve a model locally and test different versions to suit your needs. I just discovered another tool that's in early development that can help you parse the documents (exams) to output a different file per student. It's called AutoNL: https://www.reddit.com/r/LocalLLaMA/comments/1b3xfbc/small_benchmark_gpt4_vs_opencodeinterpreter_67b/ <--- see this thread for a benchmark of a local model against GPT-4.
Even then; be conscious about vulnerabilities: https://www.reddit.com/r/LocalLLaMA/s/bmtP8e1iXM
All types of insect infestations including bedbugs, cockroaches, termites are controlled with 100% guarantee.
Website: https://www.tigerpestcontrol.com.bd/
I will check out Ollama to dig deeper and see how it would work for my project.
Again, thank you very much.
All types of insect infestations including bedbugs, cockroaches, termites are controlled with 100% guarantee.
Website: https://www.tigerpestcontrol.com.bd/