app.py

import streamlit as st
from utils import naive_bayes_predict



st.sidebar.subheader('About the App')
st.sidebar.write('Text Classification App with Streamlit using a trained Naive Bayes model')
st.sidebar.write("This is just a small text classification app. Don't fret if the prediction is not correct or if it is not what you expected, the model is not perfect.")
st.sidebar.write("There is no provision for neutral text, yet...")


#start the user interface
st.title("Text Classification App")
st.write("Type in your text below and don't forget to press the enter button before clicking/pressing the 'Classify' button")

my_text = st.text_input("Enter the text you want to classify", "Change this...", max_chars=100, key='to_classify')

if st.button('Classify', key='classify_button'):
    p = naive_bayes_predict(my_text)
    if p > 0:
        # the predicted class is 1
        st.write(f"Your input text is positive and the score is {p}")
    else:
        # otherwise the predicted class is 0
        st.write(f"Your input text is negative and the score is {p}")