1/29/25, 10:12 AM SMS Spam Prediction

In [1]: import numpy as np # linear algebra

import pandas as pd # data processing, CSV file I/O (e.g pd.read_csv)
import nltk

In [9]: df = pd.read_csv(r"D:\CAPSTONE PROJECT_DEPLOYMENT\11. CAPSTONE PROJECT_DEPLOYMEN

df.head()

Out[9]: Unnamed: Unnamed: Unnamed:

v1 v2
2 3 4

Go until jurong point, crazy.. Available

0 ham NaN NaN NaN
only ...

1 ham Ok lar... Joking wif u oni... NaN NaN NaN

Free entry in 2 a wkly comp to win FA

2 spam NaN NaN NaN
Cup fina...

U dun say so early hor... U c already

3 ham NaN NaN NaN
then say...

Nah I don't think he goes to usf, he

4 ham NaN NaN NaN
lives aro...

In [11]: df= df.drop(["Unnamed: 2", "Unnamed: 3", "Unnamed: 4"], axis=1)

df= df.rename(columns={'v1':'lable', 'v2':'sms'})
df.head(6)

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 1/18

1/29/25, 10:12 AM SMS Spam Prediction

Out[11]: lable sms

0 ham Go until jurong point, crazy.. Available only ...

1 ham Ok lar... Joking wif u oni...

2 spam Free entry in 2 a wkly comp to win FA Cup fina...

3 ham U dun say so early hor... U c already then say...

4 ham Nah I don't think he goes to usf, he lives aro...

5 spam FreeMsg Hey there darling it's been 3 week's n...

In [13]: print(len(df))

5572

In [15]: df.lable.value_counts()

Out[15]: lable
ham 4825
spam 747
Name: count, dtype: int64

In [17]: df.duplicated().sum()

Out[17]: 403

In [19]: df=df.drop_duplicates(keep='first')

In [21]: df.duplicated().sum()

Out[21]: 0

In [23]: df.describe()

Out[23]: lable sms

count 5169 5169

unique 2 5169

top ham Go until jurong point, crazy.. Available only ...

freq 4516 1

In [25]: df.loc[:,'label'] = df.lable.map({'ham':0, 'spam':1})

print(df.shape)
df.head()

(5169, 3)

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 2/18

1/29/25, 10:12 AM SMS Spam Prediction

C:\Users\chitt\AppData\Local\Temp\ipykernel_12664\2601625900.py:1: SettingWithCop
yWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stabl

e/user_guide/indexing.html#returning-a-view-versus-a-copy
df.loc[:,'label'] = df.lable.map({'ham':0, 'spam':1})
Out[25]: lable sms label

0 ham Go until jurong point, crazy.. Available only ... 0

1 ham Ok lar... Joking wif u oni... 0

2 spam Free entry in 2 a wkly comp to win FA Cup fina... 1

3 ham U dun say so early hor... U c already then say... 0

4 ham Nah I don't think he goes to usf, he lives aro... 0

In [29]: df['num_characters'] = df['sms'].apply(len)

C:\Users\chitt\AppData\Local\Temp\ipykernel_12664\2372758055.py:1: SettingWithCop
yWarning:
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stabl

e/user_guide/indexing.html#returning-a-view-versus-a-copy
df['num_characters'] = df['sms'].apply(len)

In [33]: import matplotlib.pyplot as plt

import seaborn as sns
%matplotlib inline
df['num_characters'].plot(bins=50, kind='hist')

Out[33]: <Axes: ylabel='Frequency'>

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 3/18

1/29/25, 10:12 AM SMS Spam Prediction

In [35]: df.hist(column='num_characters', by='label', bins=50, figsize=(10,4))

Out[35]: array([<Axes: title={'center': '0'}>, <Axes: title={'center': '1'}>],

dtype=object)

In [39]: plt.pie(df['lable'].value_counts(), labels=['ham', 'spam'],autopct="%0.2f")

plt.show()

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 4/18

1/29/25, 10:12 AM SMS Spam Prediction

In [41]: # num of words

import nltk
nltk.download('punkt')
df['num_words'] = df['sms'].apply(lambda x:len(nltk.word_tokenize(x)))

[nltk_data] Downloading package punkt to

[nltk_data] C:\Users\chitt\AppData\Roaming\nltk_data...
[nltk_data] Package punkt is already up-to-date!

In [43]: df

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 5/18

1/29/25, 10:12 AM SMS Spam Prediction

Out[43]: lable sms label num_characters num_words

Go until jurong point, crazy.. Available

0 ham 0 111 24
only ...

1 ham Ok lar... Joking wif u oni... 0 29 8

Free entry in 2 a wkly comp to win FA

2 spam 1 155 37
Cup fina...

U dun say so early hor... U c already

3 ham 0 49 13
then say...

Nah I don't think he goes to usf, he

4 ham 0 61 15
lives aro...

... ... ... ... ... ...

This is the 2nd time we have tried 2

5567 spam 1 161 35
contact u...

5568 ham Will Ì_ b going to esplanade fr home? 0 37 9

Pity, * was in mood for that. So...any

5569 ham 0 57 15
other s...

The guy did some bitching but I acted

5570 ham 0 125 27
like i'd...

5571 ham Rofl. Its true to its name 0 26 7

5169 rows × 5 columns

In [45]: df['num_sentences'] = df['sms'].apply(lambda x:len(nltk.sent_tokenize(x)))

In [47]: df[['num_characters','num_words','num_sentences']].describe()

Out[47]: num_characters num_words num_sentences

count 5169.000000 5169.000000 5169.000000

mean 78.977945 18.455794 1.965564

std 58.236293 13.324758 1.448541

min 2.000000 1.000000 1.000000

25% 36.000000 9.000000 1.000000

50% 60.000000 15.000000 1.000000

75% 117.000000 26.000000 2.000000

max 910.000000 220.000000 38.000000

In [49]: df[df['label'] == 0][['num_characters', 'num_words', 'num_sentences']].describe(

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 6/18

1/29/25, 10:12 AM SMS Spam Prediction

Out[49]: num_characters num_words num_sentences

count 4516.000000 4516.000000 4516.000000

mean 70.459256 17.123782 1.820195

std 56.358207 13.493970 1.383657

min 2.000000 1.000000 1.000000

25% 34.000000 8.000000 1.000000

50% 52.000000 13.000000 1.000000

75% 90.000000 22.000000 2.000000

max 910.000000 220.000000 38.000000

In [51]: import seaborn as sns

In [57]: plt.figure(figsize=(12,6))
sns.histplot(df[df['label'] == 0]['num_characters'])
sns.histplot(df[df['label'] == 1]['num_characters'],color='red')

Out[57]: <Axes: xlabel='num_characters', ylabel='Count'>

In [61]: plt.figure(figsize=(12,6))
sns.histplot(df[df['label'] == 0]['num_words'])
sns.histplot(df[df['label'] == 1]['num_words'],color='red')

Out[61]: <Axes: xlabel='num_words', ylabel='Count'>

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 7/18

1/29/25, 10:12 AM SMS Spam Prediction

In [63]: sns.pairplot(df,hue='label')

Out[63]: <seaborn.axisgrid.PairGrid at 0x16dcdeaae70>

In [65]: numeric_df = df.select_dtypes(include=['number'])

# Plot correlation heatmap

sns.heatmap(numeric_df.corr(), annot=True)

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 8/18

1/29/25, 10:12 AM SMS Spam Prediction

Out[65]: <Axes: >

Data Preprocessing
Lower case

Tokenization

Removing special characters

Removing stop words and punctuation

stemming

In [69]: df['sms'][10]

Out[69]: "I'm gonna be home soon and i don't want to talk about this stuff anymore tonig
ht, k? I've cried enough today."

In [71]: from nltk.stem.porter import PorterStemmer

ps = PorterStemmer()
ps.stem('loving')

Out[71]: 'love'

In [73]: from nltk.corpus import stopwords

import nltk
nltk.download('stopwords')

[nltk_data] Downloading package stopwords to

[nltk_data] C:\Users\chitt\AppData\Roaming\nltk_data...
[nltk_data] Package stopwords is already up-to-date!

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 9/18

1/29/25, 10:12 AM SMS Spam Prediction

Out[73]: True

In [77]: import nltk

import string
from nltk.corpus import stopwords
from nltk.stem import PorterStemmer

# Initialize Porter Stemmer

ps = PorterStemmer()

# Define the function to transform the text

def transform_text(text):
text = text.lower()
text = nltk.word_tokenize(text)

y = []
for i in text:
if i.isalnum():
y.append(i)

text = y[:]
y.clear()

for i in text:
if i not in stopwords.words('english') and i not in string.punctuation:
y.append(i)

text = y[:]
y.clear()

for i in text:
y.append(ps.stem(i))

return " ".join(y)

# Apply the text transformation

df['transformed_text'] = df['sms'].apply(transform_text)

In [79]: df

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 10/18

1/29/25, 10:12 AM SMS Spam Prediction

Out[79]: lable sms label num_characters num_words num_sentences transformed

Go until
jurong
go jurong
point,
0 ham 0 111 24 2 crazi avail b
crazy..
great w
Available
only ...

Ok lar...
ok lar joke
1 ham Joking wif 0 29 8 2
u oni...

Free entry
in 2 a
free entri
wkly
2 spam 1 155 37 2 comp win f
comp to
final tk
win FA
Cup fina...

U dun say
so early
u dun say
3 ham hor... U c 0 49 13 1
hor u c alrea
already
then say...

Nah I
don't
nah think go
think he
4 ham 0 61 15 1 live a
goes to
th
usf, he
lives aro...

... ... ... ... ... ... ...

This is the
2nd time
2nd tim
we have
5567 spam 1 161 35 4 contact u p
tried 2
prize 2 cla
contact
u...

Will Ì_ b
going to b go espla
5568 ham 0 37 9 1
esplanade
fr home?

Pity, * was
in mood
5569 ham for that. 0 57 15 2 piti mood su
So...any
other s...

The guy
did some
guy bitch a
bitching
5570 ham 0 125 27 1 interes
but I
someth els
acted like
i'd...

5571 ham Rofl. Its 0 26 7 2 rofl true

true to its

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 11/18

1/29/25, 10:12 AM SMS Spam Prediction

lable sms label num_characters num_words num_sentences transformed

name

5169 rows × 7 columns

In [81]: from wordcloud import WordCloud

wc = WordCloud(width=500, height=500,min_font_size=10,background_color='white')

In [83]: spam_wc = wc.generate(df[df['label'] == 1]['transformed_text'].str.cat(sep=" "))

In [90]: plt.figure(figsize=(15,6))
plt.imshow(spam_wc)

Out[90]: <matplotlib.image.AxesImage at 0x16dd27406e0>

In [92]: df.head()

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 12/18

1/29/25, 10:12 AM SMS Spam Prediction

Out[92]: lable sms label num_characters num_words num_sentences transformed_tex

Go until
jurong
go jurong poin
point,
0 ham 0 111 24 2 crazi avail bugi n
crazy..
great world..
Available
only ...

Ok lar...
Joking ok lar joke wif u
1 ham 0 29 8 2
wif u on
oni...

Free
entry in
2 a wkly free entri 2 wkl
2 spam comp to 1 155 37 2 comp win fa cup
win FA final tkt 21..
Cup
fina...

U dun
say so
early
u dun say earl
3 ham hor... U c 0 49 13 1
hor u c alreadi say
already
then
say...

Nah I
don't
think he nah think goe us
4 ham goes to 0 61 15 1 live around
usf, he though
lives
aro...

In [94]: spam_corpus = []
for msg in df[df['label'] == 1]['transformed_text'].tolist():
for word in msg.split():
spam_corpus.append(word)

In [96]: len(spam_corpus)

Out[96]: 9939

In [100… from collections import Counter

sns.barplot(x=pd.DataFrame(Counter(spam_corpus).most_common(30))[0],
y=pd.DataFrame(Counter(spam_corpus).most_common(30))[1])
plt.xticks(rotation='vertical')
plt.show()

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 13/18

1/29/25, 10:12 AM SMS Spam Prediction

In [102… ham_corpus = []
for msg in df[df['label'] == 0]['transformed_text'].tolist():
for word in msg.split():
ham_corpus.append(word)

In [106… from collections import Counter

sns.barplot(x=pd.DataFrame(Counter(ham_corpus).most_common(30))[0], y=pd.DataFra
plt.xticks(rotation='vertical')
plt.show()

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 14/18

1/29/25, 10:12 AM SMS Spam Prediction

In [108… # Text Vectorization

# Using Bag of Words
df.head()

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 15/18

1/29/25, 10:12 AM SMS Spam Prediction

Out[108… lable sms label num_characters num_words num_sentences transformed_tex

Go until
jurong
go jurong poin
point,
0 ham 0 111 24 2 crazi avail bugi n
crazy..
great world..
Available
only ...

Ok lar...
Joking ok lar joke wif u
1 ham 0 29 8 2
wif u on
oni...

Free
entry in
2 a wkly free entri 2 wkl
2 spam comp to 1 155 37 2 comp win fa cup
win FA final tkt 21..
Cup
fina...

U dun
say so
early
u dun say earl
3 ham hor... U c 0 49 13 1
hor u c alreadi say
already
then
say...

Nah I
don't
think he nah think goe us
4 ham goes to 0 61 15 1 live around
usf, he though
lives
aro...

Model Building
In [111… from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer
cv = CountVectorizer()
tfidf = TfidfVectorizer(max_features=3000)

In [113… X = tfidf.fit_transform(df['transformed_text']).toarray()

In [115… X.shape

Out[115… (5169, 3000)

In [117… y = df['label'].values

In [119… from sklearn.model_selection import train_test_split

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 16/18

1/29/25, 10:12 AM SMS Spam Prediction

In [121… X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.2,random_state=

In [123… from sklearn.naive_bayes import GaussianNB,MultinomialNB,BernoulliNB

from sklearn.metrics import accuracy_score,confusion_matrix,precision_score

In [125… gnb = GaussianNB

mnb = MultinomialNB
bnb = BernoulliNB

In [127… from sklearn.preprocessing import LabelEncoder

# Initialize LabelEncoder
label_encoder = LabelEncoder()

# fit and trensform label in y_train

y_train_encoded = label_encoder.fit_transform(y_train)

# Transform labels in y_test

y_test_encoded = label_encoder.transform(y_test)

In [137… from sklearn.naive_bayes import GaussianNB

from sklearn.metrics import accuracy_score, confusion_matrix, precision_score

# Initialize GaussianNB
gnb = GaussianNB()

# Fit the model

gnb.fit(X_train, y_train_encoded)

# Predict on the test set

y_pred1 = gnb.predict(X_test)

# Evaluate the model

print("Accuracy:", accuracy_score(y_test_encoded, y_pred1))
print("Confusion Matrix:\n", confusion_matrix(y_test_encoded, y_pred1))
print("Precision:", precision_score(y_test_encoded, y_pred1))

Accuracy: 0.8694390715667312
Confusion Matrix:
[[788 108]
[ 27 111]]
Precision: 0.5068493150684932

In [141… from sklearn.naive_bayes import MultinomialNB

from sklearn.metrics import accuracy_score, confusion_matrix, precision_score

# Initialize Multinomial Naive Bayes

mnb = MultinomialNB()

# Fit the model

mnb.fit(X_train, y_train_encoded)

# Predict on the test set

y_pred1 = mnb.predict(X_test)

# Evaluate the model

print("Accuracy:", accuracy_score(y_test_encoded, y_pred1))

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 17/18

1/29/25, 10:12 AM SMS Spam Prediction

print("Confusion Matrix:\n", confusion_matrix(y_test_encoded, y_pred1))

print("Precision:", precision_score(y_test_encoded, y_pred1))

Accuracy: 0.9709864603481625
Confusion Matrix:
[[896 0]
[ 30 108]]
Precision: 1.0

In [143… from sklearn.naive_bayes import BernoulliNB

from sklearn.metrics import accuracy_score, confusion_matrix, precision_score

# Initialize Bernoulli Naive Bayes

bnb = BernoulliNB()

# Fit the model

bnb.fit(X_train, y_train_encoded)

# Predict on the test set

y_pred1 = bnb.predict(X_test)

# Evaluate the model

print("Accuracy:", accuracy_score(y_test_encoded, y_pred1))
print("Confusion Matrix:\n", confusion_matrix(y_test_encoded, y_pred1))
print("Precision:", precision_score(y_test_encoded, y_pred1))

Accuracy: 0.9835589941972921
Confusion Matrix:
[[895 1]
[ 16 122]]
Precision: 0.991869918699187

In [147… from sklearn.ensemble import ExtraTreesClassifier

etc = ExtraTreesClassifier(n_estimators=50, random_state=2)

In [149… from sklearn.svm import SVC

svc = SVC(kernel='sigmoid', gamma=1.0, probability=True)
mnb = MultinomialNB()
etc = ExtraTreesClassifier(n_estimators=50, random_state=2)

from sklearn.ensemble import VotingClassifier

In [151… voting = VotingClassifier(estimators=[('svm', svc), ('nb', mnb), ('et', etc)],vo

In [154… y_pred1 = voting.predict(X_test)

print("Accuracy",accuracy_score(y_test_encoded,y_pred1))
print("Precision",precision_score(y_test_encoded,y_pred1))

Accuracy 0.9816247582205029
Precision 0.9917355371900827

Completed
In [ ]:

file:///C:/Users/chitt/Downloads/SMS Spam Prediction .html 18/18