Skip to main content

​
Overview

The Streamlit web interface provides an intuitive, interactive way to classify news articles without writing code. Users can paste article text directly into a browser and get instant classification results.

​
Installation

Before running the app, ensure you have Streamlit installed: 
pip install streamlit

​
Running the Application

1

Train the Model First

Make sure you have trained the model and generated the .pkl files:
python fake_news_ia.py
This creates:
  • modelo_fake_news.pkl - The trained Logistic Regression model
  • vectorizer_tfidf.pkl - The TF-IDF vectorizer
2

Launch the Streamlit App

Start the web interface:
streamlit run app.py
The app will automatically open in your default browser at http://localhost:8501
3

Use the Interface

  1. Paste or type a news article into the text area
  2. Click the “Clasificar Noticia” button
  3. View the classification result

​
Application Architecture

The Streamlit app (app.py:1-78) consists of three main components:

​
1. Model Loading

When the app starts, it loads the trained models:
app.py:8-19
try:
    # Load trained model and vectorizer
    modelo = joblib.load('modelo_fake_news.pkl')
    vectorizer = joblib.load('vectorizer_tfidf.pkl')
    stop_words = set(stopwords.words("english"))
    
    print("Modelos y Vectorizador cargados exitosamente.")
except FileNotFoundError:
    st.error("Error: Archivos de modelo o vectorizador (.pkl) no encontrados.")
    st.error("Asegúrate de ejecutar 'fake_news_ia.py' primero.")
    sys.exit()
If the .pkl files are missing, the app displays an error message and exits gracefully.

​
2. Text Preprocessing

The same limpiar_texto function used in training:
app.py:24-39
def limpiar_texto(texto):
    # 1. Remove metadata/source
    texto = re.sub(r'([A-Z\s]+)\s*\((REUTERS|AP|AFP)\)\s*\-\s*', '', str(texto), flags=re.IGNORECASE)
    
    # 2. Convert to lowercase
    texto = str(texto).lower()
    
    # 3. Remove punctuation, numbers, special characters
    texto = re.sub(r'[^a-z\s]', '', texto) 
    
    # 4. Tokenize with split()
    tokens = texto.split() 

    # 5. Filter stopwords and single-letter tokens
    tokens = [t for t in tokens if t not in stop_words and len(t) > 1]
    return " ".join(tokens)

​
3. User Interface

The interactive classification workflow:
app.py:43-78
st.title("📰 Detector de Noticias Falsas (IA)")
st.markdown("---")

st.header("Ingresa la noticia a clasificar:")

# Text input area
noticia_input = st.text_area(
    "Pega el texto de la noticia aquí:",
    height=200,
    placeholder="Ej: The European Union formally approved a new trade agreement..."
)

# Classification button
if st.button("Clasificar Noticia"):
    if noticia_input:
        with st.spinner('Clasificando...'):
            # 1. Clean the text
            noticia_limpia = limpiar_texto(noticia_input)
            
            # 2. Vectorize using trained vectorizer
            noticia_vec = vectorizer.transform([noticia_limpia])
            
            # 3. Make prediction
            prediccion = modelo.predict(noticia_vec)[0]
            
            # 4. Display result
            st.markdown("### Resultado de la Clasificación:")
            
            if prediccion == 'real':
                st.success(f"✅ La noticia es clasificada como **{prediccion.upper()}**")
                st.balloons()  # Celebration animation
            else:
                st.error(f"❌ La noticia es clasificada como **{prediccion.upper()}**")
                
    else:
        st.warning("Por favor, pega el texto de una noticia para clasificar.")

​
User Interface Components

​
Text Input Area

The main input component where users paste news articles:
  • Height: 200px for comfortable reading
  • Placeholder: Example text to guide users
  • Variable: noticia_input stores the user’s text

​
Classification Button

Triggers the prediction pipeline when clicked:
  1. Validates input is not empty
  2. Shows spinner during processing
  3. Processes text through the full pipeline
  4. Displays formatted results

​
Results Display

Two different result presentations: 
if prediccion == 'real':
    st.success(f"✅ La noticia es clasificada como **{prediccion.upper()}**")
    st.balloons()  # Shows celebration animation

​
Example Usage

​
Testing with Sample Articles

Try these example news articles: Real News Example: 
The European Union formally approved a new trade agreement with Canada on Thursday following a vote in the European Parliament in Brussels. Officials said the agreement is expected to strengthen economic cooperation and reduce tariffs on industrial goods over the next five years.
Fake News Example: 
A secret meeting was held at the UN headquarters where delegates voted to replace all sugary drinks with green juice to boost the global population by 500 years.

​
Prediction Workflow

1

User Input

User pastes news article text into the text area
2

Text Cleaning

limpiar_texto() removes metadata, converts to lowercase, removes punctuation, and filters stopwords
3

Vectorization

vectorizer.transform() converts cleaned text into TF-IDF numerical features
4

Classification

modelo.predict() uses Logistic Regression to classify as ‘real’ or ‘fake’
5

Result Display

Green success message with balloons for real news, red error message for fake news

​
Error Handling

The app handles several error conditions: Missing Model Files: 
st.error("Error: Archivos de modelo o vectorizador (.pkl) no encontrados.")
st.error("Asegúrate de ejecutar 'fake_news_ia.py' primero.")
Empty Input: 
st.warning("Por favor, pega el texto de una noticia para clasificar.")

​
Customization

You can customize the app by modifying app.py:
  • Change the title and headers
  • Adjust text area height
  • Modify result display styling
  • Add additional input validation
  • Include confidence scores in results

​
Next Steps

Build docs developers (and LLMs) love

Get started for freeTalk to us