vanipandit27/sarcasmdetection

Developed a sarcasm detection system using Random Forest and SVM algorithms on a dataset of 1 million entries, achieving an accuracy of 85%. • Implemented advanced data pre-processing techniques, including text cleaning, tokenization, and feature engineering, resulting in a 20% improvement in accuracy for sarcasm detection models

SarcasmDetection

Developed a sarcasm detection system using Random Forest and SVM algorithms on a dataset of 1 million entries, achieving an accuracy of 85%. • Implemented advanced data pre-processing techniques, including text cleaning, tokenization, and feature engineering, resulting in a 20% improvement in accuracy for sarcasm detection models

import pandas as pd from sklearn.feature_extraction.text import CountVectorizer from sklearn.ensemble import RandomForestClassifier from sklearn.svm import SVC from sklearn.metrics import accuracy_score, confusion_matrix, classification_report from sklearn.model_selection import train_test_split

df = pd.read_csv('/content/saarcasm.csv') df.head()

df.columns

missing_values = df.isnull().sum() print(missing_values)

df.fillna(df.mean(), inplace=True)

import nltk import re nltk.download('stopwords') stemmer = nltk.SnowballStemmer("english") from nltk.corpus import stopwords import string stopword=set(stopwords.words('english'))

def clean(comment): comment = str(comment).lower() comment = re.sub('[.?]', '', comment) comment = re.sub('https?://\S+|www.\S+', '', comment) comment = re.sub('<.?>+', '', comment) comment = re.sub('[%s]' % re.escape(string.punctuation), '', comment) comment = re.sub('\n', '', comment) comment = re.sub('\w*\d\w*', '', comment) comment = [word for word in comment.split(' ') if word not in stopword] comment=" ".join(comment) comment = [stemmer.stem(word) for word in comment.split(' ')] comment=" ".join(comment) return comment df["comment"] = df["comment"].apply(clean)

import matplotlib.pyplot as plt from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator comment = " ".join(i for i in df.comment) stopwords = set(STOPWORDS) wordcloud = WordCloud(stopwords=stopwords, background_color="white").generate(comment) plt.figure( figsize=(15,10)) plt.imshow(wordcloud, interpolation='bilinear') plt.axis("off") plt.show()

df["label"] = df["label"].map({0: "Not Sarcastic", 1: "Sarcastic"}) df = df[["comment", "label"]] print(df.head())

X = df['comment'] y = df['label']

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

vectorizer = CountVectorizer() X_train = vectorizer.fit_transform(X_train) X_test = vectorizer.transform(X_test)

rfc = RandomForestClassifier(n_estimators=100, random_state=62) rfc.fit(X_train, y_train) y_pred_rfc = rfc.predict(X_test) print("Random Forest Classifier") print("Accuracy:", accuracy_score(y_test, y_pred_rfc)) print("Confusion Matrix:") print(confusion_matrix(y_test, y_pred_rfc)) print("Classification Report:") print(classification_report(y_test, y_pred_rfc))

svm = SVC(kernel='linear', random_state=42) svm.fit(X_train, y_train) y_pred_svm = svm.predict(X_test) print("SVM Classifier") print("Accuracy:", accuracy_score(y_test, y_pred_svm)) print("Confusion Matrix:") print(confusion_matrix(y_test, y_pred_svm)) print("Classification Report:") print(classification_report(y_test, y_pred_svm))

import matplotlib.pyplot as plt

rfc_score = accuracy_score(y_test, y_pred_rfc) svm_score = accuracy_score(y_test, y_pred_svm)

fig, ax = plt.subplots() ax.bar(['Random Forest', 'SVM'], [rfc_score, svm_score]) ax.set_ylabel('Accuracy') ax.set_title('Comparison of Random Forest and SVM on Sarcasm Dataset') plt.show()

new_comment = "Hey it's okay they were rained on by debris and body bits coz Trump screwed him on the deal...right guys?" new_comment_vectorized = vectorizer.transform([new_comment]) rfc_predicted_label = rfc.predict(new_comment_vectorized) svm_predicted_label = svm.predict(new_comment_vectorized) print("Random Forest predicted label:", rfc_predicted_label) print("SVM predicted label:", svm_predicted_label)

new_comment = "Loving this West Ham implosion" new_comment_vectorized = vectorizer.transform([new_comment]) rfc_predicted_label = rfc.predict(new_comment_vectorized) svm_predicted_label = svm.predict(new_comment_vectorized) print("Random Forest predicted label:", rfc_predicted_label) print("SVM predicted label:", svm_predicted_label)