Question Generation through neural model

Project Title

Question Generation through neural model

Project Area of Specialization Computer ScienceProject Summary

The field of artificial intelligence (AI) in Pakistan is evolving rapidly and poised to grow significantly over the coming decade. AI is a technology that is creating new opportunities in fields like education. To help students learn simple concepts without a teacher, AI can be beneficial.  Our software can generate questions using Deep Learning. The objective of Question Generation is to generate relevant and fluent questions, according to a given passage. This not only allows students to evaluate themselves but can also save costs by eliminating the need for assessment tutors. We have implemented an attention mechanism in this system to generate questions that are closest to what can be generated by a person. This software can be used by our educational dept to generate question paper as manual construction of questions is a complex process that requires training, experience, and resources. To reduce the expenses associated with manual construction of questions and to satisfy the need for a continuous supply of new questions. We aim to generate intelligent and relevant questions from the provided text using this system. Through this computerized application, we can reduce the task of an educator. Using this system, the student will be no longer dependent on tutor/mentor which save the cost and student can also self-assets themselves by using this system.

The field of artificial intelligence (AI) in Pakistan is evolving rapidly and poised to grow significantly over the coming decade. AI is a technology that is creating new opportunities in fields like education. To help students learn simple concepts without a teacher, AI can be beneficial.  Our software can generate questions using Deep Learning. The objective of Question Generation is to generate relevant and fluent questions, according to a given passage. This not only allows students to evaluate themselves but can also save costs by eliminating the need for assessment tutors. We have implemented an attention mechanism in this system to generate questions that are closest to what can be generated by a person. This software can be used by our educational dept to generate question paper as manual construction of questions is a complex process that requires training, experience, and resources. To reduce the expenses associated with manual construction of questions and to satisfy the need for a continuous supply of new questions.

We aim to generate intelligent and relevant questions from the provided text using this system. Through this computerized application, we can reduce the task of an educator. Using this system, the student will be no longer dependent on tutor/mentor which save the cost and student can also self-assets themselves by using this system.

Project Objectives

1. This system can be used to facilitate the Education institutions like universities, colleges, etc. to generate questions for the quizzes, assessments, or even FAQs
2. Self-assessments of the students without depending on mentor. 
3. reduce the task of professors by automating the process of generating questions.
4. The system will generate the question closest to which the experienced professor will generate.

Project Implementation Method

• Step 1: Gather all the requirements like dataset needed for the implementation            
• Step 2: Generate the front-end using reactJS. 
• Step 3: Implement the Seq2Seq model with attention mechanism.
• Step 4: Trained the model.
• Step 5: Integrating the trained model using API to the front-end.   
• Step 6: Testing the complete system.
   

Benefits of the Project

The system will be designed in view to help the educational institution, students, and tutors. 

Benefits for Educational institutions:
1.It can reduce the expenses of the education boards associated with the manual construction of questions by automating the question generation process eliminating the need for professional and trained professors.
2.It can be used for quizzes, assessments, or even FAQs in education institutions.

Benefits for Students and tutors:
3.The system can benefit the teachers by saving their time to make quizzes, test papers, or examination questions.
4.Student can also get benefit from this system by focusing more on the preparation of paper rather than making questions for self-assessments
5.Student will be no longer dependent on tutor/mentor which save the cost and student can also self-assets themselves by using this system. 
 

Technical Details of Final Deliverable

Deliverables:

Website integrated with python backend. For the development of the website, we have used reactJS. We have integrated this website using API to the backend which is developed on python.

Technical Details

• Seq2Seq Model with an attention mechanism
• Bi-direction LSTM
• The highly trained model generates accurate questions

Final Deliverable of the Project Software SystemCore Industry ITOther Industries Education Core Technology Artificial Intelligence(AI)Other Technologies Big DataSustainable Development Goals Quality EducationRequired Resources

Elapsed time in (days or weeks or month or quarter) since start of the project	Milestone	Deliverable
Month 1	Requirement gathering	• Analyzing dataset to use. • Analyze System Objectives, Constraints, and Scope • Analyze Available Resources, Benefits, and Feasibility • Prepare and Submit Preliminary Project Proposal
Month 2	Dataset Preprocessing	• Analyzing dataset • Preprocessing SQuAD dataset • Design block diagram of the complete Model • Review of Finalized ER Diagram
Month 3	System development	• Implementing Encoder of Seq2Seq model • Implementing Decoder of Seq2Seq model • Designing frontend on react native • Training the model • Exporting the trained model • connecting the export model with API • Implementing API in frontend • Finalize the development part • Code Implementation, • Review the system
Month 4	Integration	• Integrating the trained model with react frontend using API
Month 5	Testing	• Unit Testing • Module Testing • Integration Testing • System Testing