Students Virtual Assistant

Project Title

Students Virtual Assistant

Project Area of Specialization Artificial IntelligenceProject Summary

Currently, all the students have to take of all their academics related queries to the program coordinators or faculty members. Program coordinators holding an important office have a lot of responsibilities which makes it difficult for them to address every student timely. Moreover, if student have any academic issue and want to consult their concerned faculty member, they face time and availability issues. On the other hand, considering students currently pursuing their final year project, they do not have any source of information regarding which faculty members are affiliated with which particular field. Moreover, there is no source of communication with alumni who have done projects related to field of interest of students.

Realizing the marked difference between student-teacher ratio and critical importance of provision of timely and effective solution to student queries, an automated system that could handle and respond to student queries as effectively as human is much needed. This would encourage students to ask queries they might have in a targeted domain without any hesitation. It would also help them to take effective decision based on guidance obtained from chatbot and perform their academic activities in efficient manner.

Taking benefit of the property of chatbot to generate responses that have high degree of resemblance to human responses, we will develop such a system that will aid students by providing effective and timely solutions to their queries regarding courses, faculty as well as final year project. Our chatbot named, Student's Virtual Assistant will accept text-based questions of students as input. These questions will be analyzed and processed using Natural Language Processing (NLP) techniques which will tokenize student queries, identify the intention of student's query and generate appropriate response accordingly. This will save the time of both faculty and students and will provide an efficient platform to students where they can seek effective solutions to their queries.

Project Objectives

This project is intended to:

1. Identify the intention of student's query

2. Generate appropriate responses to students based on the intent of query

3. Continuously learn from chat with student and improve the quality of response

4. Provide effective and appropriate responses to students remotely, 24/7

Project Implementation Method

In this project, first of all dataset is collected by gathering most frequently asked questions by university students from three main domain i.e. courses, faculty and fyp and then recording appropriate responses for thr queries by consulting the concerned official. After collection of dataset, a pre-processing algorithm is developed using Natural Language Processing (NLP) techniques, which will pre-process query asked by the student during chat Student's Virtual Assistant.

Then a machine learning model is developed by implementing SVM (Support Vector Machine) algorithm. This model is trained on the basis of student's query dataset. The processed query generated by pre-processing alogrithm is then fed to this trained model which indentifies the intention (pre-defined class label) of student's query.

After identifying the intention of student's query an appropriate response is generated from dataset using data mining techniques.

Student's Virtual Assistant continuously learns during chat with student based on student feedback and enhance the quality of its response. If student does not obtain a satisfiable response to his/her query, he/she is provided an option to direct his/her query to program coordinator via email.

Benefits of the Project

Following are the key benefits of students virtual assistant:

1. Students obtain appropriate responses for their queries remotely, 24/7 without the need to contact or pay a visit to any official.

2. Humans are limited by their ability to respond to only one person at a time. Whereas, Student's Virtual Assistant can respond to multiple students at the same time

3. Student's Virtual Assistant continuously learn from its chat with student and enhance the quality of its response (thereby generating more human like responses).

4. Students can freely ask queries from Student's Virtual Assistant without any hesitation.

5. It saves time of both faculty and student by providing responses remotely.

Technical Details of Final Deliverable

The final deliverable project is a web based solution, which maintains a separate profile for each student and enables the students to enter their query from any of the three specified domain (i.e. courses, faculty and fyp). The query is fetched from web interface and is fed to pre-processing algorithm. The out of pre-processing algorithm is then fed to previously trained SVM model. This identifies the intention of student's query. Based on the intent of query an appropriate response is fetched from data source using data mining techniques and is delivered to student via interface.

Chatbot named Student's Virtual Assistant continuously learns during chat with student based on the feedback obtained from student and enhance the quality of its response over time.

If student is unable to obtain a satisfiable response from chatbot he/she can direct his/her query to program coordinator via email.

Final Deliverable of the Project Software SystemCore Industry EducationOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Quality EducationRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	60850
8GB Memory module	Equipment	2	5000	10000
Smart Phone	Equipment	1	35000	35000
Website domain and hosting	Equipment	1	8750	8750
Printing documents	Miscellaneous	4	1400	5600
Project CDs	Miscellaneous	3	500	1500