app.py

# utilities
import re
import pickle
import numpy as np
import pandas as pd
# plotting
import seaborn as sns
from wordcloud import WordCloud
import matplotlib.pyplot as plt
# nltk
from nltk.stem import WordNetLemmatizer
# sklearn
from sklearn.svm import LinearSVC
from sklearn.naive_bayes import BernoulliNB
from sklearn.linear_model import LogisticRegression

from sklearn.model_selection import train_test_split
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics import confusion_matrix, classification_report

from datasets import load_dataset

dataset = load_dataset('Twitter_Emoticon_Analysis_NLP/training.1600000.processed.noemoticon.csv')

DATASET_COLUMNS  = ["sentiment", "ids", "date", "flag", "user", "text"]
DATASET_ENCODING = "ISO-8859-1"
dataset = pd.read_csv('training.1600000.processed.noemoticon.csv',
                      encoding=DATASET_ENCODING , names=DATASET_COLUMNS)

# Removing the unnecessary columns.
dataset = dataset[['sentiment','text']]
# Replacing the values to ease understanding.
dataset['sentiment'] = dataset['sentiment'].replace(4,1)

# Storing data in lists.
text, sentiment = list(dataset['text']), list(dataset['sentiment'])

def preprocess(textdata):
    processedText = []
    
    # Create Lemmatizer and Stemmer.
    wordLemm = WordNetLemmatizer()
    
    # Defining regex patterns.
    urlPattern        = r"((http://)[^ ]*|(https://)[^ ]*|( www\.)[^ ]*)"
    userPattern       = '@[^\s]+'
    alphaPattern      = "[^a-zA-Z0-9]"
    sequencePattern   = r"(.)\1\1+"
    seqReplacePattern = r"\1\1"
    
    for tweet in textdata:
        tweet = tweet.lower()
        
        # Replace all URls with 'URL'
        tweet = re.sub(urlPattern,' URL',tweet)
        # Replace all emojis.
        for emoji in emojis.keys():
            tweet = tweet.replace(emoji, "EMOJI" + emojis[emoji])        
        # Replace @USERNAME to 'USER'.
        tweet = re.sub(userPattern,' USER', tweet)        
        # Replace all non alphabets.
        tweet = re.sub(alphaPattern, " ", tweet)
        # Replace 3 or more consecutive letters by 2 letter.
        tweet = re.sub(sequencePattern, seqReplacePattern, tweet)

        tweetwords = ''
        for word in tweet.split():
            # Checking if the word is a stopword.
            #if word not in stopwordlist:
            if len(word)>1:
                # Lemmatizing the word.
                word = wordLemm.lemmatize(word)
                tweetwords += (word+' ')
            
        processedText.append(tweetwords)
        
    return processedText
    
def preprocess(textdata):
    processedText = []
    
    # Create Lemmatizer and Stemmer.
    wordLemm = WordNetLemmatizer()
    
    # Defining regex patterns.
    urlPattern        = r"((http://)[^ ]*|(https://)[^ ]*|( www\.)[^ ]*)"
    userPattern       = '@[^\s]+'
    alphaPattern      = "[^a-zA-Z0-9]"
    sequencePattern   = r"(.)\1\1+"
    seqReplacePattern = r"\1\1"
    
    for tweet in textdata:
        tweet = tweet.lower()
        
        # Replace all URls with 'URL'
        tweet = re.sub(urlPattern,' URL',tweet)
        # Replace all emojis.
        for emoji in emojis.keys():
            tweet = tweet.replace(emoji, "EMOJI" + emojis[emoji])        
        # Replace @USERNAME to 'USER'.
        tweet = re.sub(userPattern,' USER', tweet)        
        # Replace all non alphabets.
        tweet = re.sub(alphaPattern, " ", tweet)
        # Replace 3 or more consecutive letters by 2 letter.
        tweet = re.sub(sequencePattern, seqReplacePattern, tweet)

        tweetwords = ''
        for word in tweet.split():
            # Checking if the word is a stopword.
            #if word not in stopwordlist:
            if len(word)>1:
                # Lemmatizing the word.
                word = wordLemm.lemmatize(word)
                tweetwords += (word+' ')
            
        processedText.append(tweetwords)
        
    return processedText

import gradio as gr
import nltk
from nltk.sentiment.vader import SentimentIntensityAnalyzer

# Download required NLTK resources
nltk.download('vader_lexicon')

# Load the pre-trained sentiment intensity analyzer
sia = SentimentIntensityAnalyzer()

def get_sentiment(tweet):
    # Preprocess the tweet
    processed_tweet = preprocess([tweet])
    
    # Get the sentiment score using VADER sentiment analyzer
    sentiment_score = sia.polarity_scores(processed_tweet[0])
    
    # Determine the sentiment label based on the compound score
    compound_score = sentiment_score['compound']
    if compound_score >= 0.05:
        sentiment = 'Positive'
    elif compound_score <= -0.05:
        sentiment = 'Negative'
    else:
        sentiment = 'Neutral'
    
    return sentiment

# Create a Gradio interface
iface = gr.Interface(
    fn=get_sentiment,
    inputs='text',
    outputs='text',
    title='Tweet Sentiment Analyzer',
    description='Enter a tweet with text or emoticon or both, and get the sentiment prediction.',
    examples=[['I love this movie!', 'This weather is terrible.']],
    theme='Soft'
    
)

# Launch the interface
iface.launch(share = True)