Building a Streamlit Web Application for NLP Tasks: Sentiment Analysis, Entity Extraction, and Text Summarization

The article introduces the broad field of web/app development for machine‑learning models and explains why Streamlit, a simple Python API, is ideal for quickly building front‑end interfaces that showcase NLP services without requiring extensive full‑stack expertise.

First, install Streamlit (e.g., pip3 install streamlit ) and import it in your script.

Use Streamlit’s title, header, and subheader functions to create a clear UI:

st.title("Natural Language Processing Web Application Example")
st.subheader("What type of NLP service would you like to use?")

Add a dropdown menu for task selection with st.selectbox :

option = st.selectbox('NLP Service', ('Sentiment Analysis', 'Entity Extraction', 'Text Summarization'))

Provide a text input area for the user’s document:

st.subheader("Enter the text you'd like to analyze.")
text = st.text_input('Enter text')

Sentiment Analysis : Tokenize the input into sentences, compute polarity with TextBlob, and plot the scores using st.line_chart . Also display overall polarity and subjectivity.

from textblob import TextBlob
sents = sent_tokenize(text)
entireText = TextBlob(text)
sentScores = []
for sent in sents:
    sentScores.append(TextBlob(sent).sentiment[0])
st.line_chart(sentScores)
sentimentTotal = entireText.sentiment
st.write("The sentiment of the overall text below.")
st.write(sentimentTotal)

Named Entity Recognition (NER) : Load spaCy’s English model, extract entities and their labels, then organize them by type using a helper function.

import spacy
from spacy import displacy
nlp = spacy.load('en_core_web_sm')
entities = []
entityLabels = []
doc = nlp(text)
for ent in doc.ents:
    entities.append(ent.text)
    entityLabels.append(ent.label_)
entDict = dict(zip(entities, entityLabels))

def entRecognizer(entDict, typeEnt):
    return [ent for ent in entDict if entDict[ent] == typeEnt]

entOrg = entRecognizer(entDict, "ORG")
entPerson = entRecognizer(entDict, "PERSON")
# ... other types
st.write("Organization Entities: " + str(entOrg))
st.write("Personal Entities: " + str(entPerson))

Text Summarization : Use Gensim’s summarize function to generate a concise summary of the input text, displaying it with st.subheader and st.write .

from gensim.summarization.summarizer import summarize
summWords = summarize(text)
st.subheader("Summary")
st.write(summWords)

The full script combines all imports, UI elements, and conditional logic for the three NLP services, providing a compact yet functional example that can be run with a single streamlit run command.

# Complete example
import streamlit as st
import pandas as pd
import matplotlib.pyplot as plt
from textblob import TextBlob
from nltk.tokenize import sent_tokenize
from gensim.summarization.summarizer import summarize
import spacy
nlp = spacy.load('en_core_web_sm')

st.title("Natural Language Processing Web Application Example")
option = st.selectbox('NLP Service', ('Sentiment Analysis', 'Entity Extraction', 'Text Summarization'))
st.subheader("Enter the text you'd like to analyze.")
text = st.text_input('Enter text')

if option == 'Sentiment Analysis':
    # sentiment analysis code (as above)
    pass
elif option == 'Entity Extraction':
    # NER code (as above)
    pass
else:
    # summarization code (as above)
    pass