Predict If The Text Is Spam Or Not In NLP

texts = [
    "Win cash now!!! Click this link",
    "Congratulations, you have won a prize",
    "Hey, are we meeting today?",
    "Please call me when you are free",
    "Limited offer! Buy now and get 50% off",
    "Let's have lunch tomorrow"
]

labels = ["spam","spam","ham","ham","spam","ham"]

# tokenizer
def spacy_tokenizer(text):
    doc = nlp(text)
    tokens = [] 
    for token in doc:
        if not token.is_stop:  
            tokens.append(token.lemma_)
    return tokens

# Initialize the Vectorizer
#Learns the vocabulary from the input data (fit). (the goal is to build a dictionary of all unique words and calculate the importance of those words)
#Transforms the text data into a feature matrix (transform).
vectorizer = TfidfVectorizer(tokenizer=spacy_tokenizer, lowercase=True)
X = vectorizer.fit_transform(texts)

# Initialize the Model
model = LinearSVC()
model.fit(X, labels)

# predict
new_message = ["Congratulations! You win free cash now"]
X_new = vectorizer.transform(new_message)
prediction = model.predict(X_new)
print("Prediction:", prediction[0])

from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.svm import LinearSVC
import spacy

# Load the English model
nlp = spacy.load("en_core_web_sm")

texts = [
    "Win cash now!!! Click this link",
    "Congratulations, you have won a prize",
    "Hey, are we meeting today?",
    "Please call me when you are free",
    "Limited offer! Buy now and get 50% off",
    "Let's have lunch tomorrow"
]

labels = ["spam","spam","ham","ham","spam","ham"]

# tokenizer
def spacy_tokenizer(text):
    doc = nlp(text)
    tokens = [] 
    for token in doc:
        if not token.is_stop:  
            tokens.append(token.lemma_)
    return tokens


# Initialize the Vectorizer
#Learns the vocabulary from the input data (fit). (the goal is to build a dictionary of all unique words and calculate the importance of those words)
#Transforms the text data into a feature matrix (transform).
vectorizer = TfidfVectorizer(tokenizer=spacy_tokenizer, lowercase=True)
X = vectorizer.fit_transform(texts)


# Initialize the Model
model = LinearSVC()
model.fit(X, labels)

# predict
new_message = ["Congratulations! You win free cash now"]
X_new = vectorizer.transform(new_message)
prediction = model.predict(X_new)
print("Prediction:", prediction[0])

Query/Feedback