SAMRAT ASHOK TECHNOLOGICAL INSTITUTE VIDISHA

COMPUTER SCIENCE & ENGINEERING DEPARTMENT

Major project presentation

Submitted By : Guided By :
Shivanshu Dwivedi (0108CS191113) Dr. Sunil Joshi
Shivendra Singh Tomar (0108CS191114)
Shreyansh Gour (0108CS191115)
Shruti Arya (0108CS191116)
INDEX

➢ Problem statement
➢ Proposed Solution
➢ Minimum system requirements
➢ Modules required
➢ Methodology
➢ Conclusion
 PROBLEM STATEMENT
● In today’s world , everything is digital. In the education system work is very lengthy and time consuming and
also requires extra manpower. Every campus needs a campus guide. At the start of any academic year,newly
enrolled students have a long list of queries. Colleges usually have a bustling amount of students, all lining up
outside the administration office to get various information.

● Each student comes in with a different query, be it regarding their course, class timings, fees, etc. Now in this
process, the faculty has to type out the student’s name or register number to find the necessary information
and then pass it onto the student.

● College staff waste a lot of their time answering the same questions year after year for new students. A
college guide is essential for freshmen to learn more about their college. A college inquiry chatbot could be
the perfect solution for that issue.

● Traditionally, students could get clarification from anyone who has been on campus for quite a while, a senior
student, a resident associate, or a professor. But with particular issues, they need help a lot of the time and
don't know who to ask. And those are the situations where chatbots are becoming critical for college inquiry.
 PROBLEM STATEMENT

● The sharing of inconsistent information can lead to a negative student experience, even if the staff member truly
believed they were relaying the correct information over the course of the student interaction. But again, without
the proper checks and balances in place to ensure that all administrations have uniform access to the same
information—for example, if there is an update in any information but if the administration is unaware of the
updated information they he delivered wrong information.

● One of the major problems with users is to be accessible and responsive whenever they need most.
Unfortunately, for many businesses, it doesn’t make sense to staff a 24/7 customer support team just to ensure
that those customer questions or requests that happen in the middle of the night get addressed by a live agent.
This is where chatbots can really make a positive impact.

● The average university website hosts a lot of information, and optimising it in real time based on changing
student needs is unrealistic. Most universities undertake website development as a standalone project, and plan
for periodic updates at suitable times during the year. A number of student preferences and issues may get
overlooked in this process, and only some of them may get addressed depending on the timing of the next
website update.
 PROPOSED SOLUTION

● A Chat-bot is a software application used to conduct an online chat conversation via text or text-to speech,
instead of providing direct contact with a live human agent.Designed to convincingly simulate the way a
human would behave as a conversational partner. In the proposed system, we presented a chatbot that
generates a dynamic response for online user's queries.

● Since it is responding dynamic response the desired answer will be generated for the user. The proposed
system used machine learning algorithms to learn the Chatbot by experiencing various user's responses and
requests.

● Nowadays chat-bot is started to becoming so robust because Artificial Intelligence aids the human touch in
every conversation, chat-bot understand the user's query, and trigger an accurate response.
PROPOSED SOLUTION

Our ChatBot will have following features :

➢ 24/7 support
➢ Automation
➢ More Accurate
➢ Less Human interaction
➢ Reduce Cost
➢ Fast Response
➢ Updated
MINIMUM SYSTEM REQUIREMENT

● Any operating system capable of running a web browser.

● NLTK (Natural Language Toolkit): NLTK is a leading platform for building Python programs to work with human
language data.
● TensorFlow: TensorFlow is a free and open-source software library for dataflow and differentiable programming
across a range of tasks.
● Numpy: NumPy is a library for the Python programming language, adding support for large, multi-dimensional
arrays and matrices, along with a large collection of high-level mathematical functions to operate on these
arrays.
● ChatterBot 1.0.5 : ChatterBot is a machine-learning based conversational dialog engine build in Python
which makes it possible to generate responses based on collections of known conversations.
● Flask: Flask is a lightweight WSGI web application framework.
● random : This module is used to generate random responses from the Chatbot.
● json : To read from json file
● pickle : To save data into files.
METHODOLOGY

● Firstly, a dataset has to be created containing the required questions and answers about the college
information. The chatbot requires at least 500,000 conversations for it to understand the semantics and
grammar. The data set has to be modified and a minimum of 1000-1500 conversations have to be added in
that data set regarding various college related enquiries.

● The chatbot is then trained on this data set for generally 100 epochs with the aim of reducing the loss error
rate and increasing the accuracy.In an ideal situation, the accuracy would be 1.000. and loss error would be
0.000. For execution, When the user runs the chatbot,the chatbot dialog box opens and user can enter its
query and submit it.

● Once the query is submitted, the query is sent for data preprocessing. The data is cleaned of various
symbols, numbers and stop words are removed. Every word is mapped to its root word and put in the same
category for easy revival. Higher form of English such as the words, I’d, I’ll and so on are converted to their
base form.
METHODOLOGY
● This cleaned data is converted into a bag of words model where every word is given a unique integer id to
be used later during the answer construction process.

● The checkpoint model i.e., the trained data is loaded into the chatbot and a prediction/thought vector is
created.This context/thought vector that generates the data based upon the previous conversations.

● The prediction vector predicts the values with the help of the neural network and the output is generated
word by word which is displayed to the user in the chatbot dialog box.
METHODOLOGY
 CONCLUSION

Artificial Intelligence is the fastest growing technology everywhere in the world. With the help of Artificial
Intelligent and Knowledgeable database.we can make the transformation in the pattern matching and virtual
assistance.This system is developing chatbot based on machine learning tensorflow so with the combination of
Artificial Intelligent Knowledgeable database and virtual assistance.we can developed such chat bot which will
make a conversion between human and machine and will satisfy the question raised by user.

The main objectives of the project were to develop an algorithm that will identify the answers associated with
user submitted queries. A database is developed to store all related data and to develop an internet interface.
The web interface developed had one parts, and that is for the administrator. A background research happened,
including a summary of the conversation procedure and any relevant chat bots available. A database system was
designed, that stores information regarding questions, answers, keywords, logs and feedback messages.