Research Paper: A.I.
Based Virtual Personal Assistant
Abhay [abhaypmms@gmail.com]
Sooraj Pratap [soorajpratap05@gmail.com]
Abhishek Gautam [abhishekgautam730249@gmail.com]
Ritesh Yadav [yadavritesh829@gmail.com]
(Department of C.S.E. & I.T. Veer Bahadur Singh Purvanchal University Jaunpur 211001)
Keyword -: AI, Python, Voice Assistant, Virtual Assistant, Speech Recognition, Personal Assistant, Desktop Assistant
ABSTRACT
As we know Python is an emerging language so it becomes easy to write a script for Voice Assistant in Python.
The instructions for the assistant can be handled as per the requirement of user. Speech recognition is the
process of converting speech into text. This is commonly used in voice assistants like Alexa, Siri, etc. In
Python there is an API called Speech Recognition which allows us to convert speech into text. It was an
interesting task to make my own assistant. It became easier to send emails without typing any word, searching
on Google without opening the browser, and performing many other daily tasks like playing music, opening
your favourite IDE with the help of a single voice command. In the current scenario, advancement in
technologies is such that they can perform any task with same effectiveness or can say more effectively than
us.
Functionalities of this project include:
1. It can send emails.
2. It can read PDF.
3. It can send text on WhatsApp.
4. It can open command prompt, your favourite IDE, notepad etc.
5. It can play music.
6. It can do Wikipedia searches for you.
7. It can open websites like Google, YouTube, etc., in a web browser.
8. It can give weather forecast.
9. It can give desktop reminders of your choice.
10. It can have some basic conversation.
Now the basic question arises in mind that how it is an AI? The virtual assistant that I have created is like if it
is not an A.I, but it is the output of a bundle of the statement. But fundamentally, the mail purpose of A.I
machines is that it can perform human tasks with the same efficiency or even more efficiently than humans.
1. INTRODUCTION 2. FEATURES
Artificial Intelligence when used with machines, it ➢ Tasks: A task is a personal or work-
shows us the capability of thinking like humans. In related assignment you want to track
this, a computer system is designed in such a way through completion. A task can occur once
that typically requires interaction from human. As or repeat (a recurring task). A recurring
we know Python is an emerging language so it task can repeat at regular intervals or repeat
becomes easy to write a script for Voice Assistant based on the date you mark the task
in Python. The instructions for the assistant can be complete. For example, you might want to
handled as per the requirement of user. Speech send a status report to your manager on the
recognition is the Alexa, Siri, etc. In Python there last Friday of every month, and get a
is an API called Speech Recognition which allows haircut when one month has passed since
us to convert speech into text. It was an interesting your last haircut. Recurring tasks are added
task to make my own assistant. It became easier to one at a time to the task list. When you
send emails without typing any word, searching on mark one occurrence of the task complete,
Google without opening the browser, and the next occurrence appears in the list.
performing many other daily tasks like playing Users can also create Task Requests. A task
music, opening your favourite IDE with the help of request enables the user to assign tasks to
a single voice command. In the current scenario, the other people, and to receive task
advancement in technologies is such that they can requests from others. When someone
perform any task with same effectiveness or can assigns a task, that person gives up
say more effectively than us. By making this ownership of the task (unless the task is
project, I realized that the concept of AI in every declined). Anyone who assigns a task can
field is decreasing human effort and saving time. keep an updated copy in their task list and
As the voice assistant is using Artificial receive status reports for the task.
Intelligence hence the result that it is providing are Associated with task requests is a task list.
highly accurate and efficient. The assistant can help ➢ Internet Applications: The Personal
to reduce human effort and consumes time while Assistant allows personnel to access,
performing any task, they removed the concept of customize, and engage the internet to help
typing completely and behave as another individual them source information ranging from
to whom we are talking and asking to perform task. weather, directions, and schedules, to stock
The assistant is no less than a human assistant but performance, competitive data, and news.
we can say that this is more effective and efficient All using simple, conversational voice
to perform any task. The libraries and packages commands, e.g. trip management, airline
used to make this assistant focuses on the time reservation and hotel reservations. The
complexities and reduces time. The functionalities convergence of the richness of the internet
include, it can send emails, it can read PDF, it can and the accessibility and mobility of the
send text on WhatsApp, it can open command phone is now forming a vast new network
prompt, your favourite IDE, notepad etc., It can - a Voice Web - where Internet content can
play music, it can do Wikipedia searches for you, it be accessed from any phone, anywhere,
can open websites like Google, YouTube, etc., in a using human voice. A voice portal can be
web browser, it can give weather forecast, It can defined as "speech-enabled access to Web
give desktop reminders of your choice. It can have based information." In other words, a voice
some basic conversation. Tools and technologies portal provides telephone users with a
used are PyCharm IDE for making this project, and natural-language interface to access and
I created all Python files in PyCharm. Along with retrieve Web content. An Internet browser
this I used following modules and libraries in my can provide Web access from a computer
project. pyttsx3, Speech Recognition, Datetime. but not from a telephone. A voice portal is
a way to do that. browser.
3. SYSTEM ARCHITECTURE 4. METHODOLOGY
➢ Take Command: The function is used to
take the command as input through
microphone of user and returns the output Speech Recognition
API calls
System call
as string. Voice input
Module
➢ Wish Me: This function greets the user
Python Backend
according to the time like Good Morning, System calls
System call
Good Afternoon and Good Evening.
Text to Speech
➢ Task Execution: This is the function which Speech Output
Extracting Data
contains all the necessary task execution
definition like send Email, pdf reader, news, 1) Python: Python is an OOPs (Object
and many conditions in if condition like Oriented Programming) based, high-level,
“open google,” “open notepad,” “search on interpreted programming language. It is a
Wikipedia”,” play music” and “open robust, highly useful language focused on
command prompt” etc. rapid application development (RAD).
Python helps the easy writing and execution
of codes. Python can implement the same
User
logic with as much as 1/5th of code as
Voice Voice compared to other OOPs languages. Python
Input Output provides a huge list of benefits to all. The
usage of Python is such that it cannot be
limited to only one activity. Its growing
Microphone Speaker popularity has allowed it to enter some of
the most popular and complex processes
like Artificial Intelligence (AI), Machine
Learning (ML), natural language
processing, Data science, etc. Python has a
lot of libraries for every need of this project.
For JARVIS, libraries used are speech
Speech – to Text – to recognition to recognize a voice, Pyttsx3
a for text-to speech, selenium for web
Text Conversion Speech Conversion
automation, etc. Python is reasonably
efficient. Efficiency is usually not a
problem for small examples. If your Python
code is not efficient enough, a general
Other procedure to improve it is to find out what
WWW is taking most of the time and implement
Application
just that part more efficiently in some
lower-level languages. This will result in
much less programming and more efficient
Request
code (because you will have more time to
optimize) than writing everything in a low-
Keyword
level language.
2) Que.py: Que.py is a Python framework to
Response
transform natural language questions into
queries in a database query language. It can
be easily customized to different kinds of
questions in natural language and database
queries. So, with a little coding, you can
build your system for natural language
access to your database. 6. USED LIBRARY’S
3) Pyttsx3: Pyttsx3 stands for Python Text to
➢ pyttsx3: It is a python library which
Speech. It is a cross-platform Python converts text to speech.
wrapper for text-to-speech synthesis. It is a ➢ Speech Recognition: It is a python module
Python package supporting common text- which converts speech to text.
to-speech engines on Mac OS X, Windows, ➢ pywhatkit: It is python library to send
and Linux. It works for both Python2.x and WhatsApp message at a particular time
3. versions. Its main advantage is that it with some additional features.
works offline. ➢ Datetime: This library provides us the
4) NLP and Voice Recognition: Natural actual date and time.
language processing (NLP) techniques are ➢ Wikipedia: It is a python module for
used to process and understand the voice searching anything on Wikipedia.
commands the desktop voice assistant ➢ Smtplib: Simple mail transfer protocol that
receives. This may involve tasks such as allows us to send mails and to route mails
speech recognition, language between mail servers.
understanding, intent recognition, and ➢ pyPDF2: It is a python module which can
context extraction, to accurately interpret read, split, merge any PDF.
the user's commands. ➢ Pyjokes: It is a python library which
5) SQLite: SQLite is a capable library, that contains lots of interesting jokes in it.
provides an in-process relational database ➢ Web browser: It provides interface for
for efficient storage of small-to-medium displaying web-based documents to users.
sized data sets. It supports most of the ➢ Pyautogui: It is a python library for
common features of SQL (Structured Query graphical user interface.
Language) with few exceptions. Best of all, ➢ OS: It represents Operating System related
most Python users do not need to install functionality.
anything to get started working with ➢ sys: It allows operating on the interpreter
SQLite, as the standard library in most as it provides access to the variables and
distributions’ ships with the sqlite3 module. functions that usually interact strongly
SQLite runs embedded in memory with the interpreter.
alongside your application, allowing you to ➢ Speed test: Python provides a library
easily extend SQLite with your own Python called speed test which is useful for testing
code. SQLite provides quite a few hooks, a the internet speed. It is basically a
reasonable subset of which are command-line interface for checking the
implemented by the standard library internet bandwidth. An internet speed test
database driver. is usually run to measure the travel speed
between your device and the server you
5. SOFTWARE REQUIREMENTS want to connect to, over the internet
connection you are using. It usually
➢ Operating system should be window 7 or displays the upload and download speeds
higher. as the result.
➢ The kernel version should be 3.0.16 or ➢ Decouple: Decouple helps you to organize
higher. your settings so that you can change
➢ Support of other basic applications like parameters without having to redeploy
maps, calendar, camera, web connection your app.
etc
• store parameters in .env files;
• define comprehensive default values;
• properly convert values to the correct
data type.
7. WORKING PRINCIPLE 8. REAL LIFE APPLICATION
Five Steps in Natural Language ➢ Saves time: A desktop voice
Processing are: assistant which works on the voice
command offered to it, it can do
voice searching, voice-activated
LEXICAL ANALYSIS device control and can let us
complete a set of tasks.
S ➢ Conversational interaction: It
SSYNTACTIC ANALYSIS
makes it easier to complete any task
as it automatically does it by using
the essential module or libraries of
SEMANTIC ANALYSIS Python, in a conversational
interaction way. Hence any user
when instruct any task to it, they
feel like giving task to a human
DISCLOSURE INTEGRATION assistant because of the
conversational interaction for
giving input and getting the desired
output in the form of task done.
PRAGMATIC ANALYSIS
➢ Reactive nature: The desktop
assistant is reactive which means it
know human language very well
Speech Recognition Architecture: and understand the context that is
provided by the user and gives
response in the same way, i.e.
human understandable language,
English. So, user finds its reaction
Cepstral Feature in an informed and smart way.
Extraction
➢ Multitasking: The main
O application of it can be its
MFCC Feature multitasking ability. It can ask for
continuous instruction one after
fEATURE other until the user “QUIT” it.
Gaussian Acoustic Model ➢ No Trigger phase: It asks for the
instruction and listen the response
that is given by user without
P(O/W) P(W)
needing any trigger phase and then
only executes the task.
Phone
HMM Lexicon
Likelihoods
N-gram
language model
Viterbi
decoder
9. EXPECTED OUTCOME
➢ Password -
Figure 9.1- Password
➢ Alarm -
Figure 9.2 - Alarm
➢ Google Search -
Figure 9.3 – Input Google Search
Figure 9.4 – Output Google Search
10. Conclusion
Using the NLP built-in with synthetic talent we have completed a design of intelligent virtual digital assistant
that can manage applications, replies to user queries, and additionally carries out internet searches and
converses with the human by interacting intelligently and manipulates the devices, this consists of looking in
Wikipedia, opening Google, YouTube, Facebook, MS-Word, MS-Excel, MS-PowerPoint, sending mails, login
to your Gmail through voice, lock your PC, restart your PC, shut down your PC, play music, placing alarms,
getting climate notifications, etc. Thus, with the aid of examining the current voice assistant systems, we have
come up with this proposed machine which is environment friendly and helps us to be more organized. There
are lot of enhancements to be developed in the future to address the various user needs like users with various
disabilities through recent machine learning algorithms.
11. Acknowledgment
The Lord has been faithful in granting the strength, wisdom, knowledge, and the courage needed throughout
this period of study. We thank the Almighty God and Father above, for life and success of this study. We wish
to show immense appreciation to our supervisor, Mr. Pravin Kumar Pandey who executed his duties of
supervising this work in a passionate, and an expeditious manner. We are thankful to and fortunate enough to
get constant encourage, support and guidance from HOD Dr. Sanjeev Gangwar and all Teaching staff of
C.S.E. department of Veer Bahadur Singh Purvanchal University Jaunpur which helped us in
successfully completing Research Paper and Project Work. My heartfelt gratitude goes to my parents, siblings
and loved ones for their support in diverse forms. My sincere appreciation goes to my good friends as well as
my course mates. GOD BLESS YOU ALL.
12. Reference
[1] Dr. Kshama, V. Kulhalli, Dr. Kotrappa Sirbi, Mr. Abhijit J. Patankar, “Personal Assistant with Voice
Recognition Intelligence,” International Journal of Engineering Research and Technology. vol 10, no.1, pp.
416-418, (2017).
[2] Kukade, Ruchita G. Fengse, Kiran D. Rodge, Siddhi P. Ransing, Vina M. Lomte "Virtual Personal Assistant
for the Blind,” International Journal of Computer Science and Technology (JCST), vol 9, Issue 4, PP.2251-
2253, October - December 2018.
[3] Tushar Gharge, Chintan Chitroda, Nishit Bhagat, Kathapriya Giri, “AI Smart Assistant,” International
Research Journal of Engineering and Technology (IRJET), vol: 06 Issue: 01, PP-3862-3863, January 2019.
[4] M. A. Jawale, A. B. Pawar, D. N. Kyatanavar, “Smart Python Coding through Voice Recognition,”
International Journal of Innovative Technology and Exploring Engineering (IJITEE), vol: 8 Issue-10, PP-
3283-3284, August 2019.
[5] Isha S. Dubey, Jyotsna S. Verma, Ms. Arundhati Mehendale, "An Assistive System for Visually Impaired
using Raspberry Pi," International Journal of Engineering Research & Technology (IJERT), vol 8 Issue 05,
PP-608-609, May 2019.
