Topic Modelling In NLP

from sklearn.feature_extraction.text import CountVectorizer
from sklearn.decomposition import LatentDirichletAllocation

docs = [
    "I love playing cricket and football",
    "Football and cricket are popular sports",
    "Virat Kohli scored a century in cricket",
    "Apple released a new iPhone",
    "Samsung launches new Android phone",
    "Smartphones have powerful processors"
]

cv = CountVectorizer(stop_words='english')
dtm = cv.fit_transform(docs)

lda = LatentDirichletAllocation(n_components=3, random_state=42)
lda.fit(dtm)

for i, topic in enumerate(lda.components_):
    print(f"Topic {i}:")
    for word in topic.argsort()[-5:]:
        print(" ", cv.get_feature_names_out()[word])

from sklearn.feature_extraction.text import CountVectorizer
from sklearn.decomposition import LatentDirichletAllocation
import pandas as pd

docs = [
    "I love playing cricket and football",
    "Football and cricket are popular sports",
    "Virat Kohli scored a century in cricket",
    "Apple released a new iPhone",
    "Samsung launches new Android phone",
    "Smartphones have powerful processors"
]

cv = CountVectorizer(stop_words='english')
dtm = cv.fit_transform(docs)

lda = LatentDirichletAllocation(n_components=3, random_state=42)
lda.fit(dtm)

# Create document-topic matrix
doc_topic_matrix = lda.transform(dtm)

# Convert to DataFrame
doc_topic_df = pd.DataFrame(
    doc_topic_matrix,
    columns=[f"Topic {i}" for i in range(lda.n_components)]
)

print(doc_topic_df)

Query/Feedback