app.py

import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.cluster import KMeans
import gradio as gr

# Adjust the file path to where titles.csv is located in your environment
CSV_FILE_PATH = 'titles.csv'  # Update this with the correct path to your CSV file

def topic_modeling(n_clusters):
    # Read the CSV file directly from the specified file path
    df = pd.read_csv(CSV_FILE_PATH)
    
    if 'title' not in df.columns:
        return "Error: CSV does not contain a 'title' column"
    
    titles = df['title'].astype(str).tolist()  # Convert the 'title' column to a list of strings
    vectorizer = TfidfVectorizer(stop_words='english', ngram_range=(1, 2), max_df=0.9)
    X = vectorizer.fit_transform(titles)
    
    model = KMeans(n_clusters=n_clusters, random_state=42)
    df['topic'] = model.fit_predict(X)
    
    # Prepare the output to display the top titles for each topic
    output = ""
    for i in range(n_clusters):
        output += f"\n### Topic {i}\n"
        top_titles = df[df['topic'] == i]['title'].head(5)
        output += "\n".join(f"- {t}" for t in top_titles) + "\n"
    return output

# Define the Gradio interface (no file input, just a slider for n_clusters)
iface = gr.Interface(
    fn=topic_modeling,
    inputs=[gr.Slider(2, 10, step=1, value=5, label="Number of Topics")],
    outputs="markdown",
    title="Topic Modeling App (TF-IDF + KMeans)",
    description="This app clusters titles from a CSV file into topics using TF-IDF and KMeans. No file upload needed."
)

if __name__ == "__main__":
    iface.launch()