QURU Question Guru
Instructors these days are involved in variety of tasks such as teaching, management responsibilities, student counselling and their self-growth that requires a lot of time. With these responsibilities, they have to spend extra hours in making contextual questions from content passages for exams or
2025-06-28 16:34:40 - Adil Khan
QURU Question Guru
Project Area of Specialization Artificial IntelligenceProject SummaryInstructors these days are involved in variety of tasks such as teaching, management responsibilities, student counselling and their self-growth that requires a lot of time. With these responsibilities, they have to spend extra hours in making contextual questions from content passages for exams or class quizzes. Keeping the busy scheduling in view, teachers could greatly be benefitted from an Automated Question Answering System. Therefore, this study’s objective is to develop a web based application for teachers to assess them in making questions automatically out from the context given. The study will achieve the objective by implementing Natural Language Processing (NLP) and Deep Learning techniques. It is also in the planning to address the solution using two different approaches, one way is to give simple textual format i.e. txt or doc document, for this situation our framework will essentially apply NLP and information extraction techniques to extricate the conceivable inquiries alongside their answers. Be that as it may, the other way is to include lectures slides (.PPT document), to let the framework first concentrate the catchphrases from PPTs and after that creep over the web to seek related substance. Later the equivalent (i.e., NLP and deep learning) approach will be utilized to extricate addresses answers.
Project ObjectivesFollowing are the objectives :
- The system will ease teachers in making question papers.
- Teachers can generate question papers directly from their slides.
- The system will save users time, by directly providing the question paper.
User can give input in two ways: either in shape of text i.e docx,txt files, or He can give powerpoint slides i.e .ppt file. In case of text our model trained on SQuAD Dataset with BiDAF will give a set of Question and Answers from the content provided by the user.
In other way i.e if user's input is ppt slides. Our solution will first generate keywords for each slide with respect to content and title of slide and then these keywords will be used to extract some content over the internet using web scrapping methods. After obtaining specific contextual paragraph for slides our trained model will predict question and answers from this paragraph.
Benefits of the ProjectThis project will help to measure how well students or any individual have understood the paragraph or subject which is under learning. It evluates the redaer's understanding about the paragraph as well as it will also measure how focused the reader is.
These type pf question appear in many national and international test for Graduate and Undergradute Scholars. But designing these tests require a lot of mental and physical effort if it is done by human. Our solution will save time and human effort to design the same for evaluating various traits of reader.
Technical Details of Final DeliverableFinal Deliverable is website which is designed on top of Django framework. It uses python as back end language. BiDAF is also trained using python and its key libraries. For training model high processing such GPUs are needed these GPUs are embedded at server side which increases the response time and train model in less amout of time.
The clinet will have the web interface. He can provide textual format or slides of content. Our solution will likewise process and generate question and answers for the user.
User can print or save as pdf these questions and answers.
Final Deliverable of the Project Software SystemType of Industry Education Technologies Artificial Intelligence(AI)Sustainable Development Goals Quality Education, Industry, Innovation and InfrastructureRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 79371 | |||
| EVGA GeForce GTX 1070 (GPU) | Equipment | 1 | 69379 | 69379 |
| inRobert 95mm Video Card Cooling Fan | Miscellaneous | 1 | 5522 | 5522 |
| Printing Documents | Miscellaneous | 6 | 120 | 720 |
| Final Thesis Printing | Miscellaneous | 3 | 1250 | 3750 |