philipobiorah's picture
add confidence level of prediction to display
3f74701 verified
import os
from flask import Flask, jsonify, request, render_template, send_file
import pandas as pd
import torch
from transformers import BertTokenizer, BertForSequenceClassification
from collections import Counter
import matplotlib
matplotlib.use('Agg') # Prevents GUI issues for Matplotlib
import matplotlib.pyplot as plt
import base64
from io import BytesIO
# Ensure the file exists in the current directory
BASE_DIR = os.path.dirname(os.path.abspath(__file__))
FILE_PATH = os.path.join(BASE_DIR, "Student_Feedback_Dataset__20_Rows_.csv")
# Fix Permission Issues: Set Writable Directories for Hugging Face & Matplotlib
os.environ["HF_HOME"] = "/tmp"
os.environ["TRANSFORMERS_CACHE"] = "/tmp"
os.environ["MPLCONFIGDIR"] = "/tmp"
# Create directories if they don’t exist
os.makedirs(os.environ["HF_HOME"], exist_ok=True)
os.makedirs(os.environ["TRANSFORMERS_CACHE"], exist_ok=True)
os.makedirs(os.environ["MPLCONFIGDIR"], exist_ok=True)
app = Flask(__name__)
# Load Model from Hugging Face
MODEL_NAME = "philipobiorah/bert-imdb-model"
tokenizer = BertTokenizer.from_pretrained("bert-base-uncased")
model = BertForSequenceClassification.from_pretrained(MODEL_NAME)
model.eval()
# Function to Predict Sentiment + Confidence Score
def predict_sentiment(text):
if not text.strip():
return {"sentiment": "Neutral", "confidence": 0.0} # Avoid processing empty text
inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True, max_length=512)
with torch.no_grad():
outputs = model(**inputs)
probabilities = torch.nn.functional.softmax(outputs.logits, dim=1)[0] # Convert logits to probabilities
sentiment_idx = probabilities.argmax().item() # Get predicted class (0 = Negative, 1 = Positive)
confidence = probabilities[sentiment_idx].item() * 100 # Convert to percentage
sentiment_label = "Positive" if sentiment_idx == 1 else "Negative"
return {"sentiment": sentiment_label, "confidence": round(confidence, 2)}
@app.route('/')
def upload_file():
return render_template('upload.html')
@app.route('/download-sample')
def download_sample():
if os.path.exists(FILE_PATH):
return send_file(FILE_PATH, as_attachment=True)
else:
return "Error: File not found!", 404
@app.route('/analyze_text', methods=['POST'])
def analyze_text():
text = request.form.get('text', '').strip()
if not text:
return jsonify({"error": "No text provided!"}), 400 # Return JSON error message
result = predict_sentiment(text)
return jsonify(result) # Return JSON response including confidence score
@app.route('/uploader', methods=['POST'])
def upload_file_post():
if 'file' not in request.files:
return "Error: No file uploaded!", 400
f = request.files['file']
if f.filename == '':
return "Error: No file selected!", 400
try:
data = pd.read_csv(f)
# Ensure 'review' column exists
if 'review' not in data.columns:
return "Error: CSV file must contain a 'review' column!", 400
# Predict sentiment & confidence for each review
results = data['review'].astype(str).apply(predict_sentiment)
data['sentiment'] = results.apply(lambda x: x['sentiment'])
data['confidence'] = results.apply(lambda x: f"{x['confidence']}%")
# Generate summary
sentiment_counts = data['sentiment'].value_counts().to_dict()
summary = f"Total Reviews: {len(data)}<br>" \
f"Positive: {sentiment_counts.get('Positive', 0)}<br>" \
f"Negative: {sentiment_counts.get('Negative', 0)}<br>"
# Generate sentiment plot
fig, ax = plt.subplots()
ax.bar(sentiment_counts.keys(), sentiment_counts.values(), color=['red', 'blue'])
ax.set_ylabel('Counts')
ax.set_title('Sentiment Analysis Summary')
# Save plot as an image
img = BytesIO()
plt.savefig(img, format='png', bbox_inches='tight')
img.seek(0)
plot_url = base64.b64encode(img.getvalue()).decode('utf8')
plt.close(fig)
return render_template('result.html', tables=[data.to_html(classes='data')], titles=data.columns.values, summary=summary, plot_url=plot_url)
except Exception as e:
return f"Error processing file: {str(e)}", 500
if __name__ == '__main__':
app.run(host='0.0.0.0', port=7860, debug=True)