NAME – ADITYA RAI

ROLL NO – 202110101150020
SUBJECT- AI & TA -2
TOPIC – RESERCH PAPER ON VIRTUAL ASSISTANT

TITLE:-
Virtual Assistant
ABSTRACT
This report discusses ways in which new technology could be
harnessed to create an intelligent Virtual Personal Assistant (VPA)
with a focus on userbase data. It will look at examples of intelligent
programs with natural language processing that are currently
available, with different categories of support, and examine the
potential usefulness of one specific piece of software as a VPA. This
engages the ability to communicate socially through natural language
processing, holding and analysing data within the context of the user.
It is suggested that new technologies may soon make the idea of
virtual personal assistants a reality. Experiments conducted on this
system, combined with user testing, have provided evidence that a
basic program with natural language processing algorithms in the
form of a VPA, with basic natural language processing and the ability
to function without the need for other type of human input (or
programming) may already be viable
INTRODUCTION
Gone are the days when humans depended on other humans for
help or services. The digitalization of the world made sure that
humans no need to contact anyone else to seek help, they could
depend on a far more efficient and reliable device which can take
care of their everyday needs. The computers, mobiles, laptops, etc.,
became a part of us and our daily life, It could carry out simple
calculations to complex programs to reduce monotonous work and
waste of manpower. Virtual Personal Assistant has almost become a
basic necessity in all electronic devices so as to execute the required
problems easily. More than just being a bot , VPA can make life easier
for the user in various ways. Speech recognition is one of the
relatively new integration into the VPA. But, though its moderately
efficient , it is not very helpful and are not used by the user due to its
high amount of error. Though the error percentage of the upcoming
VPAs is around 5 percent, it still is not quite up to the mark to where
it becomes a basic part of the users life. Thus the projects aim is to
build a VPA with speech recognition which has a very minimal error
percentage. Voice recognition is a complex process using advanced
concepts like neural networks and machine learning. The auditory
input is processed and a neural network with vectors for each letter
and syllable is created. This is called the data set . When a person
speaks the device compares it to this vector and the different
syllables are pulled out with which it has the highest
correspondence.
Another market driver for the electronic assistant is that the
automobile has become a mobile office, making safety an
increasingly important issue. Using voice commands instead of
touch-tones is not only a convenience, it is being perceived by
consumers as a safety necessity. The global market for unified
messaging services is expected to be a considerable share of the tele
comm applications. It is estimated that it will be worth several
billions by 2007.
While indirect revenues for the carriers will be several folds. A few
companies has started offering converging products in the VPA
direction, e.g. Conita, WildFire, VoxSurf, VoiceGeneie, and VoiceTel,
and Mitel Networks, though one or two provide solutions for mobile
carrier environment.
In summary, VPA promises to provide hands-free, eyes-free access to
the web anywhere, any time, from any phone.
FEATURES OF VPA
A. TASK
A task is a personal or work-related assignment you want to track
through completion. A task can occur once or repeat (a recurring
task). A recurring task can repeat at regular intervals or repeat based
on the date you mark the task complete. For example, you might
want to send a status report to your manager on the last Friday of
every month, and get a haircut when one month has passed since
your last haircut. Recurring tasks are added one at a time to the task
list. When you mark one occurrence of the task complete, the next
occurrence appears in the list. Users can also create Task Requests. A
task request enables the user to assign tasks to the other people, and
to receive task requests from others. When someone assigns a task,
that person gives up ownership of the task (unless the task is
declined). Anyone who assigns a task can keep an updated copy in
their task list and receive status reports for the task. Associated with
task requests is a task list. A list that includes the name of the person
who originally sent the task request plus the names of everyone who
also received the task request, reassigned the task to someone else,
and chose to keep an updated copy of the task in their task lists. VPA
provides VUI to the MS Outlook task manager
B. Internet Applications
The VPA allows personnel to access, customize, and engage the
internet to help them source information ranging from weather,
directions and schedules, to stock performance, competitive data
and news. All using simple, conversational voice commands, e.g. trip
management, airline reservation and hotel reservations. The
convergence of the richness of the internet and the accessibility and
mobility of the phone is now forming a vast new network - a Voice
Web - where Internet content can be accessed from any phone,
anywhere, using human voice. A voice portal can be defined as
"speech-enabled access to Web based information." In other words,
a voice portal provides telephone users with a natural-language
interface to access and retrieve Web content. An Internet browser
can provide Web access from a computer but not from a telephone.
A voice portal is a way to do that. The emerging Voice Web (or voice
portals) is all about distributed voice-enabled content known as voice
sites that can be accessed from any telephone using a standard voice
browser. This emerging technology creates a new model for business
applications known as V – commerce.
SYSTEM ARCHITECTURE
The system architecture of this projects shows the flow of the control
through the system. It also shows the hardware and the software
required for the execution of the program. The architecture diagram
is as follow:
HARDWARE AND SOFTWARE REQUIREMENTS.
Hardware:
A phone with a touch screen interface
 Phone Ram should be of a minimum 512 MB
.  Internet connectivity.
 The phone should have USB debugging mode for development and
testing purposes.
Software:
 Operating system should be android 4.1/win 8.1/IOS 6 or higher.
 The kernel version should be 3.0.16 or higher
.  Support of other basic applications like maps, calendar, camera,
web connection etc.
EXISTING AND PROPOSED SYSTEM Existing Model
Most of the existing projects have only used speech recognition using
neural networks. Though their systems have a moderate accuracy,
they are not for practical usage nor efficient to be of any real use
There are a few rudimentary techniques used by them
1. Context-aware computing:
Context-aware computing is a class of systems that have the ability to
sense their physical environment and adapt themselves to it
accordingly. These can be used for recognizing words spoken by
people with varying accents. It can also deduce words that may have
been misspoken.
2. MFCC:
MFCC refers to the Mel-Frequency Cepstral Coefficients. MFC (Mel-
Frequency Cepstrum) is a collection of these coefficients. It amounts
to the short-term power spectrum of a sound. These can be used to
sense variations in sound so as to recognize the various variables
required for voice recognition.
3. NLP:
Natural Language Programming is a branch of Artificial Intelligence
that deals with the interactions of computer and human languages. It
mainly focuses on how to program the computers so that they can
process the large lume of data on natural languages. This concept is
used to familiarize the computer with the various words in a
particular language and also to recognize them when spoken.
Proposed Model
Speech to text:
 A Piece of software used that converts audio to text. It doesn’t
understand just anything you might say.
Text Analyzing:
:  Converted text is just letters for computer.
 A piece of software converts text to something that is
understandable for computer
.  Computer understands the command, so Virtual Assistant like siri
convert this text to computer command
.  VPAs maps the words to functions and parameters to create a
command that computer can understand.
The major milestone that this project tries to achieve is that it tries it
increase the accuracy of the speech to text software. Meaning the
software will theoretically be able to convert any speech with slight
modulations or different accents into text with high level of accuracy
and precision needed for day to day usability of the VPA. The
software essentially combines voice recognition using neural
networks and lip movement detection using machine learning to
increase the precision of the word spoken. For people with different
accents, just voice recognition will be useless because the words they
speak will be vastly different from the actual word by the computer’s
point of view because the vectors or the values stored for that
particular word would have been gotten only based on the word
being spoken in a particular accent. So here is where lip movement
recognition comes into play. For most words, though in a different
accent, the movement of the lips remains similar enough to deduce
the word. Thus, lip movement recognition helps cutting down the
various other words which would have had the same likeliness as per
the voice recognition software.
WORKING PRINCIPLES
The working of Virtual Assistant uses following principles:
 Natural Language Processing : Natural Language Processing (NLP)
refers to AI method of communicating with an intelligent systems
using a natural language such as English.
Processing of Natural Language is required when you want an
intelligent system like robot to perform as per your instructions,
when you want to hear decision from a dialogue based clinical expert
system, etc.
Five Steps in Natural Language Processing are:
 Automatic Speech Recognition : To understand command according
to user’s input.
 Artificial Intelligence : To learn things from user and to store all
information about behaviour and relations of user. The ability of a
system to calculate, reason, perceive relationships and analogies,
learn from experience, store and retrieve information from memory,
solve problems, comprehend complex ideas, use natural language
fluently, classify, generalize, and adapt new situations
.

 Inter Process Communication : To get important information from

other software applications
RESULT OF VA

In the contemporary, fast-paced digital landscape, the

ascent of virtual assistants has brought about a
transformative shift in how we approach work.

Propelled by technological advancements and the

escalating demand for remote work opportunities,
virtual assistants have evolved into indispensable
assets for both businesses and professionals alike.

Virtual assistants have recently witnessed a surge in

popularity, primarily owing to their capacity to deliver
efficient and flexible support remotely.
As an increasing number of businesses adopt remote
work models, there has been an exponential rise in the
demand for virtual assistants. Venturing into the realm
of virtual assistance opens up a realm of services and
tasks that can be executed from virtually any corner of
the world.

CONCLUSION:

The paper describes a new emerging service for mobile user. The
Virtual Personal Assistance provides an intelligent computer
secretarial service for mobile professionals. The new service is based
on convergence of internet, speech recognition technology and mobile
technologies. The VPA minimizes the interruption of the user,
improves the utilization of his time, and provides a single point of
communication for all his messages, contacts, schedule, and source of
information. The paper proposes as well a decision structure for call
screening, as well as handling requests for meetings and appointment.
The system initially targets lawyers, doctors, sales personnel, small
offices, maintenance crews, etc. However, it is expected to become a
standard feature for millions of other users. It overcomes many of the
drawbacks in the existing solutions. It is mainly built to make a much
more efficient VPA o that they can be brought into much more
practical day to day uses. But the system has its own limitation.
Though the efficiency is high the time consumption for each task to
complete maybe higher than the other VPAs and also the complexity
of the algorithms and the concepts would make it very tough to tweak
it if needed in the future.

Reference:

[1] G. O. Young, “Synthetic structure of industrial plastics (Book

style with paper title and editor),” in Plastics, 2nd ed. vol. 3, J. Peters,
Ed. New York: McGraw-Hill, 1964, pp. 15–64.

[2] W.-K. Chen, Linear Networks and Systems (Book styl\

e).Belmont, CA: Wadsworth, 1993, pp. 123–135.
[3] H. Poor, An Introduction to Signal Detection and Estimation.
New York: Springer-Verlag, 1985, ch. 4.

[4] B. Smith, “An approach to graphs of linear forms (Unpublished

work style),” unpublished.

[5] E. H. Miller, “A note on reflector arrays (Periodical style—

Accepted for publication),” IEEE Trans. Antennas Propagat., to be
published

[6] Ardissono, L., Boella. And Lesmo, L. (2000) “A Plan-Based

AgentArchitecture for Interpreting Natural Language Dialogue”,
International Journal of Human-Computer Studies.