Analyzing 

WhatsApp Chat
 

Using 

Python
  

Libraries


Ranjan 
$ #

Bhoomi Gupta
2023 International Conference on Intelligent Systems for Communication, IoT and Security (ICISCoIS) | 979-8-3503-3583-5/23/$31.00 ©2023 IEEE | DOI: 10.1109/ICISCOIS56541.2023.10100443

 !##


Department 

of IT  !##


Department 

of IT
!
!"

Mahara "##$#

&

ja Agrasen Institute of Technology, !
!"

Maharaja "##$#

&

Agrasen Institute of Technology,

India
Delhi, 

India
Delhi, 

!

[email protected] $ ##

[email protected]



"

Vanshika Kapoor !
 
"

Deepika Bansal
 !##


Department 

of IT  !##


Department 

of IT
!
!"

Mahara "##$#

&

ja Agrasen Institute of Technology, !
!"

Maharaja "##$#

&

Agrasen Institute of Technology,
Delhi, 


India 
India
Delhi, 


%"

[email protected]  "#
[email protected]


()')/
$,

Abstract- How "#-

many "((#

messaging %%(
apps )'

there '

are #
in )

the *6C96H

Pandas >H
is 6C

an DE:C HDJG8:
A>7G6GN

open-source library B69:

made EG>B6G>AN

primarily ;DG
for
!-)$'

PlayStore, #
in "-

my $%#$#

opinion, (
is ''!+#)

irrelevant. )(
WhatsApp %%
#)

can't 

be LDG@>C<

working H>BEAN

simply 6C9
and :;;:8I>K:AN

effectively L>I=

with G:A6I>DC6A

relational DG
or I6<<:9

tagged
'%!

replaced -

by #-$#

anyone. (

These -(

days, +'-$#(

everyone's !
life %#(

depends $#
on
)(
WhatsApp. %%
Because
*(
$
)(
*(''#!-

)(
of its user-friendly UI, WhatsApp %%
(
+'-

is very
96I6
data.
#I
EGDK>9:H
6
K6G>:IN
D;
96I6
HIGJ8IJG:H
6C9
I:8=C>FJ:H

It provides a variety of data structures and techniques
%$%*!'
')
#$,
#
-$*
,#)
)$
$##)
$'
$""*#)

popular right now. When you want to connect or communicate ;DG
LDG@>C<
L>I=
I>B:
H:G>:H
6C9
CJB:G>86A
96I6
for working with time series and numerical data.

,)

with ($"$#

someone, )(
WhatsApp %%
(

is )

the '()

first "((#

messaging ('+

service ))

that '6IEADIA>7

Matplotlib >H
is 6

a BJAI> EA6I;DGB
96I6

multi-platform data K>HJ6A>O6I>DC

visualization IDDA@>I

toolkit
$"(

comes )$
to "#

mind. )(

It's $#+##)

convenient $'
for +'-$#

everyone )$
to )

chat $#
on 9:H><C:9

designed ID

to LDG@

work L>I=

with I=:

the A6G<:

large ,8>*N

SciPy HI68@

stack 6C9

and 7J>AI

built DC

on
)(
WhatsApp.%%
$

To (#

send 
a "((

message, ("%!-

simply !*#

launch )
the %%

app, (!)

select
(JB*N

NumPy 6GG6NH .+&MIG68I
>H

arrays.URLExtract is 6
a *NI=DC

Python 8A6HH

class I=6I

that 8DAA:8IH

collects
$#))
#

Contact, and )-%

type. '

There -$*

you $
go. +#

Given ))
that )(
WhatsApp %%

'+(
"$'
)#

receives more than 65 
!!$#
"(((
+'-
(#!
-
,
#

billion messages every single day, we can
:MIG68IH

(extracts) .+&H

URLs ;GDB
from EGDK>9:9

provided >CEJIinput.
BD?>

Emoji >H
is JH:9

used ;DG

for
*(

use )(
WhatsApp %%
)(

chats )$

to #!-.

analyze $#+'()$#(

conversations ,
we +

have ,)

with :BD?>
6C6ANH>H
emoji analysis.

'#(

friends, !#)(

clients, $'

or !'

large '$*%(

groups $
of %$%!

people. $
To !'#

learn #)'*#

intriguing

#()(
$*)
,$
(
"$()
)+
,$
'
$()(
)$(
,$
$

insights about who is most active, who are ghosts (those who do ##
II.
&#-+-.+
,.+/1

LITERATURE SURVEY
#$)
'%!-
'&*#)!-
)
"$()
%$%*!'
"$
not reply frequently), the most popular emoji, 
)
*(()
$*'(

the busiest hours


$

of )

the -

day, $'
or )

the "$()

most %$%*!'

popular )'"(

terms, )

chat #!-((

analysis $#
on 

a
)(
WhatsApp %%
'$*%

group )

chat $

of $!!

college ()*#)(

students (
is *)!.

utilized. (

These
#C
DG9:G
ID
HIJ9N
I=:
JH6<:
D;
0=6IHEE
In order to study the usage of WhatsApp,
6
HIJ9N
L6H
9DC:

a study was done
,$*!

would *#$*)!-

undoubtedly 
be ($"

some #()*!

insightful ($+'(

discoveries )$
to >C

in !:GB6CN

Germany ID
to ;>C9

find DJI

out I=:

the JH6<:

usage D;

of HB6GIE=DC:H

smartphones 6C9
and
$#('(

consider.This %'$ )
*((

project uses %-)$#

python #
and )(

its !''(

libraries )$
to #!-.

analyze 0=6IHEE

WhatsApp 6BDC<

among E:DEA:
people,
I=:

the HIJ9N

study >H

is 8DC9J8I:9

conducted DC
on 

241 8
)
)
)

the chat data. HB6GIE=DC:

smartphone JH:GH
users,
;DG

for 
4 L::@H
weeks.
-=:
The G:HJAI

result H=DLH

shows I=6I

that
688DJCIH
;DG
0=6IHEE
688DJCI:9
;DG

accounts for WhatsApp accounted for 1 9.83% 
 (32.
B>C

1 1 min)

Index
Terms:Chat
 
anal
 
 
WhatsApp,
ysis, Insight,  


Python D;
HB6GIE=DC:
JH6<:
of smartphone usage.
0=6IHEE
=6H
L>IC:HH:9
:MEDC:CI>6A

WhatsApp has witnessed exponential


>C8G:6H:

increase >C

in I=:

the CJB7:G

number D;
of JH:GH

users H>C8:

since >IH

its >CIGD9J8I>DC
introduction,

I.
#INTRODUCTION
!&$"'&"!
L=>8=

which B6N

may 7:

be JH:9

used ID

to G:H:6G8=

research I=:

the G>H:

rise >C

in I=:

the CJB7:G

number D;
of
JH:GH
users ..
#C
EG

I=:
6K:G6<:
BDCI=AN
JH:G
D;
0=6IHEE

In Apr 20 13 the average monthly user of WhatsApp
6I6

Data EGD8:HH>C<

processing 6C9

and 6C6ANH>H

analysis 6G:

are I=:

the ;DJC96I>DC

foundation D;
of I=>H

this
L6H

was 

200 B>AA>DC

million L=>8=

which >C

in '6G

Mar 
2020 =6H

has 7:8DB:

become 6GDJC9

around
IDDA
tool.
6I6
>H
8DCHI6CIAN
>C8G:6H>C<
6C9
>I
=6H
ADIH
D;
JH:;JA

Data is constantly increasing and it has lots of useful 2000 million.


B>AA>DC
>C;DGB6I>DC
information.
0>I=
I=:
=:AE
D;
I:8=CDAD<N
L:
=6K:
With the help of technology we have,
L:
86C
we can '6<:9
(
Maged N.
%6B:A
DJADH
9>H8JHH:H
CJB:GDJH
G:H:6G8=
I=6I

Kamel Boulos discusses numerous research that
6C6ANO:

analyze I=:
the 96I6
data.
6I6

Data 6C6ANH>H

analysis 8DCH>HIH

consists D;
of 8A:6C>C<
cleaning,
=6K:

have 7::C

been 9DC:

done ID
to 9:I:GB>C:

determine I=:
the JH:;JAC:HH

usefulness D;
of 0=6IHEE

WhatsApp
B6C>EJA6I>C<
manipulating,
EGD8:HH>C<

processing G6L
raw 96I6
data,
6C9

and :MIG68I>C<

extracting I=:

the >C
=:6AI=
6C9
=:6AI=
86G:
>C
=>H
6GI>8A:
>C
DG9:G
ID
:M6B>C:

in health and health care in his article in order to examine
G:A:K6CI

relevant >C;DGB6I>DC

information L=>8=

which 86C
can 7:

be JH:9

used ID

to I6@:

take >C;DGB:9

informed I=:
6EEA>86I>DC
D;
0=6IHEE
>C
9>K:GH:
9DB6>CH
the application of WhatsApp in diverse domains.A 
HIJ9N
7N

study by
9:8>H>DC
decision.
88DG9>C<
ID
0=6IHEE
According to WhatsApp,
I=:G:
6G:
DK:G

7>AA>DC

there are over 5 5 billion =:JC<

Cheung :I
et 6A
a!.
24

[4] ;>C9H

finds 0=6IHEE

WhatsApp >H
is K:GN

very JH:;JA

useful >C
in
B:HH6<:H

messages G:8:>K:9

received :K:GN

every 96N
day.
-=:

The INE>86A

typical JH:G

user AD<H

logs 

195 EG:K:CI>C<

preventing G:A6EH:

relapse 6BDC<

among HBD@>C<

smoking FJ>II:GH
quitters.
-=DH:

Those L=D

who
B>CJI:H

minutes 6
a L::@

week DC

on 0=6IHEE

WhatsApp 6C9
and E6GI>8>E6I:H

participates >C
in H:K:G6A

several =6K:

have HIDEE:9

stopped HBD@>C<

smoking 6G:
are 699:9

added ID
to 6

a <GDJE

group L=:G:

where I=:N

they
<GDJEH
groups.
0:
We BJHI

must HI6GI

start I=>H

this EGD?:8I

project H>C8:

since I=:G:

there >H

is 6

a IG:6HJG:

treasure :C<6<:
>C
6
9>H8JHH>DC
L>I=
DI=:G
FJ>II:GH
engage in a discussion with other quitters.

IGDK:
D;
96I6
I=:G:
>C
;GDCI
D;
JH

L:AA
trove of data there in front of us. A well-liked A>@:9
EGD<G6BB>C<

programming

A6C<J6<:
>H
*NI=DC
language is Python.
#C
 In 1 99 1 ,
!J>9D
K6C
+DHHJB
EGD9J8:9

Guido van Rossum produced CDI=:G

Another HIJ9N

study 9DC:

done 7N

by &:K:CI

Levent :I>C@6N6

Cetinkaya 24

[5] ;>C9H

finds I=6I

that
>Iit,
6C9
>I
7:86B:
6K6>A67A:
and it became available.
0=6IHEE

WhatsApp >H
is K:GN

very JH:;JA

useful >C

in >C8G:6H>C<

increasing BDI>K6I>DC

motivation 6C9

and

*NI=DC
86C
7:
JH:9
DC
6
H:GK:G
ID
8G:6I:
L:7
6EEA>86I>DCH
Python can be used on a server to create web applications.

L>AA>C<C:HH
ID
HIJ9N
>C
HIJ9:CIH
willingness to study in students.
-=:
;>C9>C<H
H=DL:9
I=6I

The findings showed that

0DG@;ADLH

Workflows B6N

may 7:
be B69:

made 7N
by JH>C<

using >I

it >C

in 8DC?JC8I>DC

conjunction L>I=

with
HIJ9:CIH

students' HJ88:HH

success =6H

has >C8G:6H:9

increased >C

in 7DI=

both I=:

the 8DCK:CI>DC6A

conventional

6EEA>86I>DCH
applications.
#I

It 86C

can A>C@

link ID

to 96I676H:

database B6C6<:B:CI

management HNHI:BH
systems.

8A6HHGDDB

classroom H:II>C<

setting 6C9

and I=:

the H:II>C<

setting L=:G:

where 0=6IHEE

WhatsApp L6H

was
utilized as a supplementary tool.

JI>A>O:9
6H
6
HJEEA:B:CI6GN
IDDA
99>I>DC6AAN
Additionally,
>I

it =6H

has ;>A:

file G:69>C<

reading 6C9

and :9>I>C<

editing 86E67>A>I>:H
capabilities.



&6G<:
96I6
H:IH
86C
7:
B6C6<:9
6C9
8DBEA:M
B6I=:B6I>86A

Large data sets can be managed and complex mathematical
EGD8:HH:H

processes 86C

can 7:

be E:G;DGB:9

performed JH>C<

using *NI=DC
Python.
#I

It B6N

may 6AHD

also 7:

be
,6B>G
,=:G>;;
>C
=>H
6GI>8A:
H=DLH
I=:
HI:E 7N HI:E
EGD8:HH

Samir Sheriff in his article shows the step-by-step process

6EEA>:9
ID

to 7J>A9

build I=:

the 0=6IHEE

WhatsApp 8=6I

chat 6C6ANO:G

analyzer JH>C<

using I=:

the ENI=DC

python
applied ID

to I=:

the 8G:6I>DC

creation D;
of HD;IL6G:

software I=6I

that >H

is G:69N

ready ;DG

for
EGD<G6B>C<
A6C<J6<:
6C9
>IH
A>7G6G>:H
programing language and its libraries.
24
[6]
EGD9J8I>DC
DG
FJ>8@
EGDIDINE>C<
production or quick prototyping.




,:K:G6A

Several *NI=DC

Python A>7G6G>:H

libraries I=6I

that 6G:

are JH:9

used >C8AJ9:
An DE:C
include: C
open­
HDJG8:
*NI=DC
IDDA@>I
86AA:9
,IG:6BA>I
6AADLH
JH:GH
ID
7J>A9

source Python toolkit called Streamlit allows users to build
6C9
H=6G:
:N: 86I8=>C<
DG><>C6A
DCA>C:
6EEA>86I>DCH
;DG
96I6

and share eye-catching, original online applications for data
6C6ANH>H

analysis 6C9

and B68=>C:

machine A:6GC>C<
learning.
DBEA:M

Complex 96I6

data 6EEH

apps 86C

can 7:

be
8G:6I:9
6C9
9:EADN:9
>C
DCAN
6
;:L
B>CJI:H
created and deployed in only a few minutes.

979-8-3503-3583-5/23/$3 1 .00 2023

979-8-3503-3583-5/23/$31.00 @2023
c IEEE 181
181

Authorized licensed use limited to: Rajiv Gandhi University of Knowledge Technologies. Downloaded on July 07,2024 at 10:43:26 UTC from IEEE Xplore. Restrictions apply.
 ###
III.
'-"))&)!1

METHODOLOGY /6
2&0"1"3&
3)&
$)"3
-&22"(&2
'1/-
3)&
%"3&
".%
3*-&


Now separate the chat messages from the date and time -



-=:
B6@>C<
D;
I=>H
6EEA>86I>DC
8DCH>HIH
D;
I=G::
B6?DG
HI:EH

The making of this application consists of three major steps:
%'

df 
pd.DataFrame({'chat_message
0% "3"1"-&:$)"3!-&22"(&
': -&22"(&2

messages,
6

a) !:II>C<
I=:
8=6I
96I6
Getting the chat data.

-&22"(&!%"3&
%"3&2;

'message_date': dates})
7

b) *G:EGD8:HH>C<
Preprocessing.

%'-&22"(&!%"3&
df['message_date 7
=
8

c) C6ANO>C<
6C9
9>HEA6N>C<
I=:
96I6

Analyzing and displaying the data
0% 3/!%"3&3*-&%'-&22"(&!%"3&
pd. to_datetime(df['message_date J,



'/1-"3%
f - 8




ormat='%d/%m/%y, %H: %M- ')

A.
&33*.(
3)&
$)"3
%"3"


Getting the chat data -
%' 1&."-&$/,4-.2:-&22"(&!%"3&%"3&;

dfrename(columns={'message _date ': 'date'},
•

• )C

On 0=6IHEE
WhatsApp,the I=:
6EEA>86I>DC

application DE:CH

opens I=:

the 8=6I

chat I=6I

that
*.0,"$&14&

inplace=True)
NDJ
L6CI
ID
:MEDGI
you want to export.

%' )&"%

dfheadO
•

• A>8@
DC
I=:
I=G::
9DIH
>8DC
DC
I=:
IDE
G><=I
D;
I=:

Click on the three dots icon on the top right of the
H8G::C
screen.


chat_message date
•

• A>8@

Click DC

on I=:

the BDG:

more DEI>DC

option >C

in I=:

the 9GDE9DLC

dropdown B:CJ

menu
6C9
I=:C
H:A:8I
I=:
MEDGI
8=6I
DEI>DC
and then select the Export chat option.
Messages an d calls are end-to-end encrypted. N... 2022·04-27 1 7:1 6:00

1 Anupam Cst created group "CHESS PARTICt PANTS-2"\n 2022-04-27 1 7:04:00

•

• =6I
L>AA
7:
G:69N
ID
:MEDGI
>C
6
;:L
H:8DC9H
>C
6

Chat will be ready to export in a few seconds in a
I:MI
;>A:
;DGB6I

text file format Madhav Sangat added youln 2022-04-27 1 7:1 6:00











53

Fig 1.
0-@1
-:0
@591
->1
?1<->-@10
2>;9
@41
/4-@
91??-31

date and time are separated from the chat message.




*G:G:FJ>H>I:H


Prerequisites:
8=6I5B:HH6<:H

chat_messages CDL

now =6K:

have I=:
the JH:G

user C6B:

name 8DCI6>C:9

contained >C

in >I

it
-D

To EGD8::9

proceed L>I=

with I=:

the EGD?:8I

project I=:

the ;DAADL>C<

following ENI=DC

python A>7G6G>:H

libraries
L=>8=
>H
H=DLC
>C
><
1 .
,:E6G6I:
I=:
JH:G
C6B:
;GDB
I=:

which is shown in Fig. Separate the user name from the
BJHI
7:
>CHI6AA:9
DC
I=:
B68=>C:
must be installed on the machine.

8=6I5B:HH6<:
8DAJBC
chat_message colunm.

,IG:6BA>I
Streamlit,
'6IEADIA>7
Matplotlib,
,:67DGC
Seaborn,
*6C96H
Pandas,
.GA:MIG68I
Urlextract,



0DG98ADJ9
Wordcloud,
:BD?>

emoji

JH:GH

24

users = []

B.
1&01/$&22*.(

Preprocessing B:HH6<:H

24

messages = []
,IG:6BA>I

Streamlit >H

is JH:9

used ID

to B6@:

make I=:

the L:7

web 6EEA>86I>DC

application ;DG

for H=DL>C<

showing ;DG
B:HH6<:
>C
9;28=6I5B:HH6<:4

for message in df{'chat_message']:
I=:
6C6ANH>H
DC
I=:
7GDLH:G



:CIGN

G: HEA>I23L304 3H
B:HH6<:

entry = re.split('([\w\W]+?):\s', message)
the analysis on the browser.
First >GHI
D;
6AA
of all,
8=6I
96I6
;GDB
I=:

chat data from the




>;
:CIGN24


JH:G
C6B:

if entry[1 :] : # user name
JH:G
>H
ID
7:
I6@:C
HD
6
;>A:
JEAD69
7JIIDC
JH>C<
HIG:6BA>I
>H

user is to be taken so a file upload button using streamlit is
JH:GH 6EE:C9:CIGN24

users.append( entry[1])
ID
7:
B69:
to be made.
B:HH6<:H 6EE:C9
".?D>C:CIGN24

messages.append(" join(entry[2:]))
40,/"%&%!'*,&
23
uploaded_file= st.2*%&#"1 '*,&!40,/"%&1)//2&
"
'*,&

sidebar.file_uploader("Choose a file ")



:AH:

else:
JH:GH 6EE:C9<GDJE5CDI>;>86I>DC

users.append('group _notification')
(DL
I=:
8=6I
96I6
>H
>C
EA6>C
I:MI
;DGB6I
>
Now the chat data is in plain text format i.e :
>C
I=:
I:MI
;>A:
in the text file,
B:HH6<:H 6EE:C9:CIGN24

messages.append( entry[OJ)
6C9

and I=6I

that >H

is 9>;;>8JAI

difficult ID

to 6C6ANO:
analyze.
,D

So I=>H

this I:MI

text ;>A:

file >H

is ID

to 7:

be 9;2JH:G4

JH:GH

df{'user'] = users
8DCK:GI:9
>CID
6
BDG:
HIGJ8IJG:9
;DGB6I
converted into a more structured format.

9;2B:HH6<:4

B:HH6<:H

df{'message'] = messages
9; 9GDE8DAJBCH28=6I5B:HH6<:4
df.drop( colunms=['chat_message'],
>CEA68:-GJ:

inplace=True)
,>C8:

Since I=>H

this EGD?:8I

project JH:H

uses ENI=DC

python I=>H

this EA6>C

plain I:MI

text C::9

need ID

to 7:

be
8DCK:GI:9
>CID
6
E6C96H
96I6;G6B:
converted into a pandas dataframe.



date user message

-=:
8=6I
;>A:
=6H
:CIG>:H
A>@:
I=>H

The chat file has entries like this: 2022-04-27 18:16:00 Madhav Sangal All participants need to report at 9 am tomorr...

  


27/04/22, 1 8:29 -
+6C?6C

Ranjan $=6

Jha: 

A 7G>:;

brief DK:GK>:L

overview L>AA

will 7:

be 2022-04-27 18:19:00 •91 96674 23691 How can we participate in 2 evenls simuHaneou ...
<>K:C
7:;DG:
I=:
B6I8=
given before the match.

2022-04-27 18:20:00 Madhav Sangal 9 am is the time for atlendance. We will �an ...



53

Fig 2.
&41
A?1>
5?
?1<->-@10
2>;9
@41
/4-@
91??-31

The user is separated from the chat message.
*123

First /'

of ",,

all 20,*3

split 3)&

the 3*-&

time ".%

and -&22"(&

message 2*.(

Using 1&(4,"1

regular


&701&22*/.2
<

expressions -
(DL
HEA>I
I=:
8=6I5B:HH6<:
8DAJBC
>CID
JH:G
6C9
B:HH6<:

Now split the chat_message colunm into user and message


JH>C<
I=:
G:<JA6G
:MEG:HH>DC
H=DLC
>C
><
2.

using the regular expression shown in Fig.
'

/0&.$)"3'*,&
f 373
1
&.$/%*.(43'

= open('chatfile.txt', 'r', encoding= 'utf-8')
,DB:

Some BDG:

more DE:G6I>DCH

operations >H
is ID

to 7:

be 9DC:

done DC

on I=>H

this 96I6;G6B:

dataframe ID

to
%"3"

' 1&"%

data = freadO
<:I

get I=:

the 96N
day,
96I:
date,
6C9

and I>B:

time =DJG B>CJI:D;

(hour,minute )of I=:

the B:HH6<:

message
0"33&1.

pattern= %:  ; %:  2}1\d{2,4},
'\d{1,2}1\d{1, ; %:  ;2%:  ;%: ;2
\s\d{1,2}:\d{2} \s­ ;GDB
I=:
96I:
8DAJBC
JH>C<
E6C96H
86C
;JGI=:G
BD9>;N
I=>H

from the date column using pandas can further modifY this
2

\s ' 96I6;G6B:

dataframe 6H
as E:G

per I=:

the G:FJ>G:B:CI

requirement D;
of I=:

the 6C6ANH>H

analysis I=6I

that L:

we
-&22"(&2
1& 20,*30"33&1.
%"3"
messages= re.split(pattern, 

data)[l:] L6CI
ID
9D
want to do.

%"3&2

1& '*.%",,0"33&1.
%"3"

dates = re.findall(pattem, data) C.

.",89*.(
".%
%*20,"8*.(
3)&
"3"

Analyzing and displaying the Data
=6I
96I6
>H
>C
6
HIGJ8IJG:9
96I6;G6B:
Chat data is in a structured dataframe,
(DL
JH>C<
9>;;:G:CI

Now using different

ENI=DC
A>7G6G>:H
;DAADL>C<
>CH><=IH
6G:
:MIG68I:9

python libraries following insights are extracted:

#C

In -DE

Top ,I6I>HI>8H

Statistics IDI6A

total B:HH6<:H

messages H:CI
sent,
-DI6A

Total LDG9H

words H:CI
sent,

 

Dates 

Messages
':9>6

Media ;>A:
file H=6G:9
shared,
6C9

and &>C@H

Links H=6G:9

shared L>AA

will 7:

be H=DLC )I=:G

shown.Other
I=6C
I=>H
'DCI=AN
I>B:A>C:
D;
I=:
<GDJE
68I>K>IN
than this Monthly timeline of the group activity,
I=:
6>AN
the Daily
I>B:A>C:
D;
I=:
<GDJE
68I>K>IN
timeline of the group activity,
'DHI
JHN
96N
Most Busy day,
'DHI
JHN

Most Busy
P  


'27/04/22, 1 8:29 -
Q
, P+6C?6C
$=6

7G>:;
DK:GK>:L

'Ranjan Jha: A brief overview
'DCI=
Month,
'DHI
68I>K:
JH:G
Most active user,
6C9
0DG9
ADJ9
D;
I=:
8=6I
L>AA

and Word Cloud of the chat will
L>AA

will 7:

be <>K:C

given 7:;DG:

before I=:

the
7:
H=DLC
be shown.
+:HJAI
D;
I=:
6C6ANH>H
>H
H=DLC
JH>C<
B6IEADIA>7

Result of the analysis is shown using matplotlib
B6I8= 3CQ

match.\n'
6C9

and HIG:6BA>I

streamlit >C
in I=:

the ;DGB

form D;
of E>8IDG>6A

pictorial G:EG:H:CI6I>DC

representation 76G

(bar

<G6E=
graph,
A>C:
<G6E=

E>:
8=6GI
6C9
I67JA6G
;DGB
line graph & pie chart) and tabular form.

182
182 2023 International
2023 International Conference
Conference on
on Intelligent
Intelligent Systems
Systems for Communication, IoT
for Communication, loT and
and Security
Security (ICISCoIS)
(ICISColS)

Authorized licensed use limited to: Rajiv Gandhi University of Knowledge Technologies. Downloaded on July 07,2024 at 10:43:26 UTC from IEEE Xplore. Restrictions apply.
 #/ Result
IV.
+:HJAI
I
)JIEJI
C6ANH>H

Output Analysis HI6GI
ID
H=>;I
ID
D;;A>C:
BD9:
6;I:G
I=:
AD8@9DLC
86JH:9
7N

start to shift to offline mode after the lockdown caused by

I=:
covid
the 8DK>9
19

virus.
K>GJH
Similarly,Fig.
,>B>A6GAN ><
5
shows
H=DLH
usJH
the
I=:
daily96>AN

ForDG
6C6ANO>C<
I=:
0=6IHEE
8=6I
96I6
;GDB
6
8A6HH
<GDJE
D;

analyzing the WhatsApp chat data from a class group of I>B:A>C:
D;
I=:
<GDJE
DK:G
I=:
E:G>D9
timeline of the group over the period.
-=:
BDHI
68I>K:
96N

The most active day
I=:
students
the HIJ9:CIH
of
D;
an
6C
Institute
#CHI>IJI:
is>H
taken
I6@:C
from
;GDB
the
I=:
author's
6JI=DGH
is>H
1
'6G8=
2021
5 March 
with
L>I=
843

messages
B:HH6<:H
sent
H:CI
as
6H
it>I
is>H
the
I=:

0=6IHEE
6EEA>86I>DC
L=D
>H
6
B:B7:G
D;
I=6I
<GDJE
WhatsApp application who is a member of that group.
-=:
The HI6GI>C<
96N
D;
I=:
I=
H:B:HI:G
starting day of the 4th semester.

8=6I
96I6
I6@:C
>H
;GDB

J<

ID

(DK

chat data taken is from 20 Aug 2020 to I I Nov 2022.


;DAADL>C<
output/visualization
following DJIEJI K>HJ6A>O6I>DC
are 6G:
obtained
D7I6>C:9
from
;GDB
the
I=:
 Most
B.
/23
428
"8
".%
/23
428
/.3)

Busy Day and Most Busy Month:
6C6ANH>H

analysis.
+'+,

Overview: 4000

Total Total Media Link 3500

Messages Words Shared Shared 3000

Ul
..
g' 2soo

20957 1 14976 1506 496

Ul
Ul

E
2000

-
Fig
53
3. 
"B1>B51C

Overview 0 1500

><
3
shows
H=DLH
how
=DL
many
B6CN
messages,
B:HH6<:H
words,
LDG9H
media,
B:9>6
and
6C9
links
A>C@H
�
Fig "E 1000
=6K:
7::C
H:CI
ID
I=:
<GDJE
7N
I=:
E6GI>8>E6CIH
have been sent to the group by the participants.
:::l

z 500

 Monthlv
A.
/.3),8
".%
%"*,8
*-&,*.&
and dailv Timeline:
..
3500
f
� I �
...i;'
0£
i;'
�
i
ia
1
i;'

� "'
)000 J Days


Fig 6. Most
53

busy day
;?@
.A?E
0-E

2500

i


2000
3500
"l
0
..., 1500
3000
3
" 1000
l!l 2500
IJ'Q
Ill
til 500 i 2000
0 �E 1500
.8
§
1000

z 500

�E � s: �
� �
1:!
�
... �c ...� l c i;' !i �

!
.. 2. ::E "'

�
g
E E
Months
::E
.!!, � � .i ...

.1! � of
Fig
53
4. 
Monthly Timeline

;:@48E
&59185:1 Months



Fig 7. Most
53

busy months
;?@
.A?E
9;:@4?



800 #C
Fig.
In ><

6C9
><

6 and Fig. 7,
#I
86C
7:
H::
I=:
7JH>:HI
96N
6C9
BDCI=

It can be see the busiest day and month
;DG
I=:
<GDJE
G:HE:8I>K:AN
for the group respectively.
-=JGH96N
>H
I=:
BDHI
68I>K:
96N

Thursday is the most active day
L=>A:
'6G8=
>H
I=:
BDHI
68I>K:
BDCI=
while March is the most active month.

I
600


&&+,8
$3*5*38
"0

C. Weeklv Activitv Map:
"l
0
...., 400


3 Friday
!l
I!
a 200 Monday 700

600
saturday

�
0 �
500

� ,., � ... ,., � c Sunday

6
N
!
N
<?
...
N
.:.
N
<?
N
N
<?
N
N
,:.
N
>!
{'!
400
0 0 0 0 0 0 0
N N N N N N N Thursday
300

200
Days
Tuesday

Fig Daily Timeline

53
5. 
-58E
&59185:1 100

Wednesday

>< 
shows
Fig.4 H=DLH
the
I=:
monthly
BDCI=AN
activity
68I>K>IN
of
D;
the
I=:
group.
<GDJE
The
-=:
graph
<G6E=

g�����������N6����·�����
" N " N � � � � � � � O M � N � O · � � � � � O

8A:6GAN
shows
clearly H=DLH
that
I=6I
March
'6G8=
2021

is>H
the
I=:
busiest
7JH>:HI
month
BDCI=
with
L>I=

.-1 .-l .-l .-1 .-1 .-l .-l .-l .-1 .-l N N N N
 
messages
3,388 B:HH6<:H
sent
H:CI
which
L=>8=
is>H
the
I=:
starting
HI6GI>C<
month
BDCI=
ofD;
the
I=:

H:B:HI:G
Gradually
semester. !G69J6AAN
the
I=:
activity
68I>K>IN
decreased
9:8G:6H:9
as 6H
the
I=:
classes
8A6HH:H

period


Fig 8. Weakly heatmap of the message
53

)1-78E
41-@9-<
;2
@41
91??-31

2023
2023International
International Conference
Conference on
onIntelligent
Intelligent Systems
Systemsffor
or Communication,
Communication,loT
IoT and
and Security
Security (ICISCoiS)
(ICISCoIS) 183
183

Authorized licensed use limited to: Rajiv Gandhi University of Knowledge Technologies. Downloaded on July 07,2024 at 10:43:26 UTC from IEEE Xplore. Restrictions apply.
 ":6IB6E

Heatmap =6H

has 6

a 8DADG

color H86A:

scale I=6I

that H=DLH

shows I=:

the CJB7:G

number D;

of I>B:A>C:
timeline,
96>AN

daily I>B:A>C:
timeline,
BDHI

most 7JHN

busy BDCI=

month 6C9

and 96N
day,
BDHI

most
B:HH6<:H

messages H:CI

sent 8DGG:HEDC9>C<

corresponding ID

to I=:

the 8DADG
color,
I=:

the CJB7:G

number D;

of 68I>K:
JH:G
active user,
L::@AN
68I>K>IN
B6E
weekly activity map,
:I8 etc.
I=6I
86C
7:
K>HJ6A>O:9

that can be visualized
B:HH6<:H
>C8G:6H:H
6H
L:
6G:
<D>C<
IDL6G9H
I=:
A><=I
8DADG
messages increases as we are going towards the li ght color.
JH>C<

using B6IEADIA>7

matplotlib A>7G6GN
library.
-=:G:

There 86C

can 7:

be BDG:

more H8DE:

scope >C

in I=:

the

8=6I

chat 6C6ANH>H

analysis DC:

one 86C

can IGN

try ID

to ;>C9

find HDB:

some DI=:G

other G:A:K6CI

relevant
><

67DK:
H=DLH
I=:
=:6IB6E
D;
B:HH6<:H
Fig. 8 above shows the heatmap of messages,
L=>8=
96N
6C9

which day and >CH><=I This insight can be used for various other purposes.

insight.
-=>H
>CH><=I
86C
7:
JH:9
;DG
K6G>DJH
DI=:G
EJGEDH:H
I>B:
I=:
B:HH6<:
=6H
7::C
H:CI
time the message has been sent.
-=:
67DK:
B6E
H=DLH
I=6I

The above map shows that

 
'

1 1 -12 AM DC'DC96N

onMonday >H
is I=:

the I>B:

time L=:C

when BDHI

most D;

of I=:

the /# References

VI.
+:;:G:C8:H
B:HH6<:H
=6K:
7::C
H:CI
ID
I=:
<GDJE
messages have been sent to the group,
I=>H
>H
I=:
I>B:
L=:C

this is the time when +[l]C.
,
Montag,
;:@-3

M-?F751C5/F
$
%->5E?7-

-/49-::

K. Blaszkiewicz, R. Sariyska, B. Lachmann,
I=:
<GDJE
>H
BDHI
68I>K:


l. :0;:1

Andone, 
B. &>1:0-258;B

5.1?

Eibes, -:0

and 
->7;C1@F
H%9->@<4;:1

the group is most active.
Trendafilov, M. A. Markowetz, "Smartphone
A?-31
5:
@41

usage in the 2 1?@
/1:@A>E
)4;
5?
-/@5B1
;:
)4-@?<<I


st century: Who is active on WhatsApp?," BMC Research   



B;8

:;

Notes, vol. 8, no. l
, 20  1 5


.

D.
/23
$3*5&
42&1

Most Active user: +[2],
#A.85?410
.E

1/5
-:0


Published b y L. Ceci and J . 2
H)4-@?<<
!A9.1>
;2
9;:@48E
-/@5B1

7 , "WhatsApp: Number of monthly active
1600 A?1>?

users 2020," I
Statista,
  
27-Jul-2022.
A8  

+[3],
M.
-918

Kamel ;A8;?

Boulos, 
D . 5A?@5:5

Giustini, -:0
and %
S . )41181>

Wheeler, H:?@-3>-9

"lnstagram -:0
and
1400 )4-@?<<
5:
1-8@4
-:0
1-8@4/->1
:
;B1>B51CI


B;8

WhatsApp in Health and Healthcare: An overview," Future Internet, val.

:;

<


8, no. 3, p. 37, 2016.  


1200
+[4],
*
&
41A:3


4-:


-5
)

4-:

Y . T. Cheung, C . H . Chan, C.-K. J. Lai, W. F . Chan,
1000
M.
#
)-:3


5
%
%
4-:
-:0
&
-9
H'?5:3
)4-@?<<
-:0

P. Wang, H. C. Li, S. S. Chan, and T.-H. Lam, ''Using WhatsApp and
� 2-/1.;;7

facebook ;:85:1
online ?;/5-8

social 3>;A<?

groups 2;>
for ?9;75:3

smoking >18-<?1

relapse <>1B1:@5;:

prevention 2;>
for >1/1:@

recent
ii BOO =A5@@1>?

<58;@
<>-39-@5/
/8A?@1>
>-:0;95F10
/;:@>;8810
@>5-8I

quitters: A pilot pragmatic cluster randomized controlled trial," Journal 


of
 

  
B;8

Medical Internet Research, vo!. 17, no. 10, 2015. 
:;

 


600 +[5],

1@5:7-E-
H&41
59<-/@
;2
)4-@?-<<
A?1
;:
?A//1??
5:
10A/-@5;:

" L. Cetinkaya, "The impact of Whatsapp use on success in education
� <>;/1??

process G
- 5:@1>:-@5;:-8

international >1B51C
review ;2
of >1?1->/4

research 5:
in ;<1:

open -:0
and 05?@>5.A@10

distributed
400
81->:5:3I

!;B  

learning," Erudit, 29-Nov-2017.
200 +,
%
%41>522
HA580
E;A>
;C:
C4-@?-<<
/4-@
-:-8EF1>I


[6] S. Sheriff, "Build your own whatsapp chat analyzer," Medium, 29-Apr- <>
 

2019.
+,)

[7]W. McKinney,/5::1E
H#-:0-?

''Pandas: 
A 2;A:0-@5;:-8

fo undational #E@4;:

Python 5.>->E

Library 2;>
for -@-

Data
i .� ·"
�
:>:
;;;
:>:
:-8E?5?
-:0
%@-@5?@5/?I

Analysis and Statistics. ••
c
:i
i E. £
>
+,

/;:-80
H1@@5:3
?@->@10
C5@4
%@>1-985@
C1.
.-?10
-<<85/-@5;:?I

[8] A. McDonald, "Getting started with Streamlit web based applications,"
E
5
� �

 -E  


Medium, 25-May-2022.
� �
!'l
+,

[9] 
D. ->/F1C?75

Karczewski, H)4E
"Why 5?
is <E@4;:

python -
a 3>1-@

great /4;5/1

choice 2;>
for 0-@-

data -:-8E?5?I

< i analysis?,"
3581
%;2@C->1
1B18;<91:@
31:/E
5:
A>;<1

Agile Software Development Agency in Europe, 19-Apr-202 <>  1


+[ 1,
%
3>-C-8
H:@>;0A/@5;:
@;
<-:0-?I
105A9
05-Jan-20
-:  2 1
.
Usm

0] S. Agrawal, "Introduction to pandas," Medium,
53

;?@
-/@5B1
A?1>

Fig 9. Most active user +[ 1 1,
] 
H . $;?1:.1>3

Rosenberg, JH)&%##

"'WHATSAPP, &$I
TEACHER?" 
- %&'!&

STUDENT

#$%#&(%

PERSPECTlVES "!

ON &$%&'!&

TEACHER-STUDENT )&%##

WHATSAPP

><

H=DLH
I=:
IDE

68I>K:
JH:GH
D;
I=:
<GDJE
L>I=
I=:
!&$&"!%
!
%"!$*
%""%K

INTERACTIONS IN SECONDARY SCHOOLS' 21-Jun-20 A:  1
8.
Fig. 9 shows the top 5 active users of the group with the

+[ 1 2],
%
%A?58-C-@5
-:0
&
%A<>5E-@:;
H":85:1
81->:5:3
@4>;A34
)4-@?<<

S . Susilawati and T . Supriyatn o, "Online leaming through WhatsApp
=><=:HI
B:HH6<:
H:CI
highest message sent.

3>;A<
5:
59<>;B5:3
81->:5:3
9;@5B-@5;:
5:
@41
1>-
-:0
<;?@
<-:0195/
/;B50

group in improving learning motivation in the era and post pandemic covid


-19,"
I
$1<;?5@;>E

Repository ;2
-A8-:-
Malik
of Maulana -857
.>-459

Ibrahim %@-@1

State ?8-95/

Islamic ':5B1>?5@E

University ;2
of

E.
'DHI
DBBDC
LDG9H
JH:9
Most Common words used:
-8-:3
0 1A:  

Malang, -Jun-2020.
+[ 1 3,
$-B5?4-:7->-

4-:A?4
(-5?-74
-:0
%>-6-:

%
H)4-@?-<<
4-@

] Ravishankara K, Dhanush, Vaisakh, and Srajan I S, "Whatsapp Chat
code
file
:-8EF1>I
:@1>:-@5;:-8
;A>:-8
;2
:35:11>5:3
$1?1->/4
-:0
B;8
(
:;

Analyzer," Intemational Journal of Engineering Research and, vol. V9, no.
onhne 05,
2020.
 


start
ma'am
+[14],
$

R. #-@58

Patil, H)4-@?<<

"WhatsApp :-8EF1>
Analyzer: )4-@?<<

WhatsApp 3>;A<
group /4-@

chat -:-8E?5?

analysis A?5:3

using
assignment #E@4;:I
:-8E@5/?
(504E-

Python," Analytics Vidhya, 30-Apr-202 <>  1

.
matns ll
tes.t +[ 1 5,

5?4>-
H)4-@?-<<
4-@
:-8EF1>
C1.
-<<

:0
@;
1:0
?@>1-985@

] H . Mishra, ''Whatsapp Chat Analyzer web app : End t o end streamlit
meet1ng
teams
<>;61/@I

proj ect," Medium,105A9
1 31/
-Dec-202  1
.
(f) viva +[ 1,

6] 

L . $

R . >/1

Arce, H)4-@?<<

''WhatsApp 3>;A<
group /4-@
chat -:-8E?5?

analysis C5@4

with <E@4;:I

python," Medium, 105A9

"'E leave
03-Nov-2020.
!;B  

0 attendance
S: roll +[ I,

7] 
A. -04-B

Jadhav, %
S . #-@58

Patil, M. 
%4-574

Shaikh, #
P. #-8

Pal, #

P. 
A. %-C-:@

Sawant, -:0
and #
P. #

P.
group
bunk A:17->

Kunekar, H)4-@?<<

"WhatsApp /4-@
chat ?1:@591:@

sentiment -:-8EF1>I

analyzer," :@1>:-@5;:-8

lnternational ;A>:-8

Journal 2;>
for
mam
$1?1->/4
5:
<<8510
%/51:/1
-:0
:35:11>5:3
&1/4:;8;3E
B;8

Research in Applied Science and Engineering Technology, vol. 1
:;
0 , no. 12, 

join
<<
G 
2022.
pp. 5 1 2-516,  


class
1;nk �
�
�������������������
0 0
+[ 1,

8 ] %

S . -45E-

14-B5;A>-8

Dahiya, 

Behavioural :-8E?5?

;4@-
-:0

A. Mohta,
Analysis ;2
of )4-@?<<

and 
A. -5:

WhatsApp /4-@?I

Jain, H&1D@

"Text /8-??525/-@5;:

chats," 2020
classification .-?10

 
5th
based
@4
:@1>:-@5;:-8

International
0 0
N ..
;:21>1:/1
;:
;99A:5/-@5;:
-:0
81/@>;:5/?
%E?@19?
%

Conference on Communication and Electronics Systems (ICCES), 2020.  


Word count
+[ 1,

9] $

R. -6<-5

Bajpai, $
R. 
K. %4->9-

Sharma, -:0
and $
R. (-?45?4@

Vashisht, H01:@525/-@5;:

"Identification ;2
o f <1-7

peak

4;A>

hour -:0
and -/@5B1

active 919.1>

member ;:
on )4-@?<<

WhatsApp 3>;A<
group A?5:3

using #E@4;:I

52@4

Common words

53

Most
;?@
;99;:
C;>0? Python," 2022 Fifth
Fig 10.
:@1>:-@5;:-8

International ;:21>1:/1

Conference ;:

on ;9<A@-@5;:-8

Computational :@188531:/1

Intelligence -:0

and

;99A:5/-@5;:
&1/4:;8;351?
&

Communication Technologies (CCICT), 2022.  


-=:
BDHI
8DBBDC
LDG9H
H:CI
7N
I=:
<GDJE
E6GI>8>E6CIH
6G:

The most common words sent by the group participants are +[20]
,

%@-:/5:
-:0

;B5/
H:
;B1>B51C
-:0
/;9<->5?;:
;2
2>11
<E@4;:

I. Stancin and A. Jovic, "An overview and comparison of free python
H=DLC

shown >C
><

in Fig. 1 0.
-=:

The H6B:

same 6C6ANH>H

analysis 86C

can 7:

be 9DC:

done ;DG

for I=:

the 85.>->51?
2;>
0-@-
95:5:3
-:0
53
-@-
:-8E?5?I

libraries for data mining and Big Data Analysis," 2019  
42nd :0
:@1>:-@5;:-8

International
>C9>K>9J6A
E6GI>8>E6CIH
D;
I=:
<GDJE
individual pa11icipants of the group.
;:B1:@5;:

Convention ;:
on :2;>9-@5;:

lnformation -:0
and ;99A:5/-@5;:

Communication &1/4:;8;3E

Technology, 81/@>;:5/?

Electronics

-:0
5/>;181/@>;:5/?
(MIPRO),
and Microelectronics #$"
2019.  


Conclusion

/ +[21],


K. %-4;;

Sahoo*, 
A. 
K. %-9-8

Samal, 
J. #>-9-:57

Pramanik, -:0
and %

S. 

K. #-:5

Pani, HD<8;>-@;>E

V.
DC8AJH>DC "Exploratory
0-@-

data -:-8E?5?

analysis A?5:3

using #E@4;:I

Python," :@1>:-@5;:-8

[ntemational ;A>:-8
Jownal ;2
of ::;B-@5B1

[JUlovative
#C
DC8AJH>DC
In Conclusion,
#I
86C
7:
H6>9
I=6I
ENI=DC
6C9
>IH
A>7G6G>:H
6G:

It can be said that python and its libraries are
&1/4:;8;3E

Technology -:0
and D<8;>5:3

Exploring :35:11>5:3

Engineering, B;8
vol. 

8, :;

no. 12, 
<<
 G  

pp. 4727-4735,
86E67A:

capable D;

of 9D>C<

doing 96I6

data 6C6ANH>H

analysis D;

of I=:

the :MEDGI:9

exported 0=6IHEE

WhatsApp
 

2019.
8=6I

chat 6C9

and K>HJ6A>O>C<

visualizing >I it.
0=6IH6EE

Whatsapp >H
is L>9:AN

widely JH:9

used 6H

as 6

a +[22],

K. 1881E

Kelley, H)4-@

''What 5?
is 0-@-

data -:-8E?5?

analysis? <>;/1??

process, @E<1?

types, 91@4;0?

methods -:0
and
8DBBJC>86I>DC
IDDA
communication tool.
-=JH
>I
=6H
6
ADI
D;
8=6I
96I6
I=6I
86C

Thus it has a lot of chat data that can &1/4:5=A1?I
%59<8581->:/;9
-:  


Techniques," Simplileam.com, 3 1-Jan-2023.
7:

be 6C6ANO:9

analyzed ID

to <:I

get HDB:

some >CH><=I
insight.
=6I

Chat 96I6

data 86C

can DCAN

onJy 7:

be

<6I=:G:9
>;
NDJ
6G:
6
E6GI>8>E6CI
>C
6
0=6IHEE
<GDJE
gathered if you are a participant in a WhatsApp group,
6C9
>I
and it

>H

is 6HHJB:9
I=6I
I=:

assumed that the H:C9:G

sender =6H

has CDI

not 9:A:I:9

deleted 6CN

any B:HH6<:H

messages
DI=:GL>H:
I=:
6C6ANH>H
L>AA
EGD9J8:
6
;6AH:
G:HJAI
otherwise the analysis will produce a false result.
I
I=:
C9

At the End
D;
I=:
LDG@
of the work,
I=:
6EE
H=DLH
I=:
9:H>G:9
G:HJAI
>:
the app shows the desired result ie.
I=:
BDCI=AN

the monthly

1 84
184 2023 International Conference
Conference on Intelligent Systems for
Systems for Communication,
Communication, IoT
loT and Security (ICISCoIS)
(ICISCoiS)

Authorized licensed use limited to: Rajiv Gandhi University of Knowledge Technologies. Downloaded on July 07,2024 at 10:43:26 UTC from IEEE Xplore. Restrictions apply.