Update app.py

app.py

@@ -165,20 +165,12 @@ def generate_performance_metrics():
 def create_interface():
     performance_metrics = generate_performance_metrics()
     with gr.Blocks() as interface:
-        with gr.Tab("Demo"):
-            gr.Markdown("Spam and Phishing Email Detection")
+        with gr.Tab(" 📨 Demo"):
+            gr.Markdown("📧🔍 Spam and Phishing Email Detection")
             gr.Markdown(
                 """
                 Welcome to the Spam and Phishing Email Detection Demo! This tool leverages DistilBERT, a lightweight yet powerful transformer model, to classify emails as ham (legitimate), spam, or phishing based on their content.
-
-                To provide a comprehensive overview of the system, the demo is divided into three key sections:
-                
-                Detection Demo – Input an email and see real-time classification results.
-                
-                Metrics Analysis – Gain insights into the performance of the model, including accuracy, precision, recall, and F1-score.
-                
-                Credits – Acknowledging the datasets, tools, and frameworks that made this project possible.
-                This project aims to enhance email security by identifying malicious messages with high accuracy, reducing the risk of scams and fraud. Feel free to explore the demo and see how AI is improving cybersecurity!
+                This project aims to enhance email security by identifying malicious messages with high accuracy, reducing the risk of scams and fraud. Feel free to explore the demo and see how AI can provide a safer environment for everyone.
                 """)    
             
     
@@ -206,7 +198,7 @@ def create_interface():
                 outputs=[result_output, confidence_output]
             )
         
-        with gr.Tab("Analysis"):
+        with gr.Tab(" 📈 Analysis"):
             gr.Markdown("## Dataset Overview")
             gr.Markdown("### Dataet Headers")
             gr.DataFrame(df)
@@ -228,9 +220,91 @@ def create_interface():
                 gr.Textbox(value=performance_metrics["recall"], label="Recall", interactive=False)
                 gr.Textbox(value=performance_metrics["f1_score"], label="F1 Score", interactive=False)
                 
-        with gr.Tab("Glossary"):
-            gr.Markdown(" ## Credits and Reference ")
+        with gr.Tab("📜 Glossary"):
+            with gr.Column():
+                with gr.Row():
+                gr.Markdown(
+                """
+                ### Label Definitions
+                - Spam: Unwanted or potentially harmful emails detected by the system.
+                - Ham: Legitimate and safe emails.
+
+                ### Evaluation Metrics
+                - Accuracy: Measures the percentage of correctly classified emails.
+                - Precision: Out of all emails classified as spam, how many were actually spam?
+                - Recall: Out of all actual spam emails, how many were identified correctly?
+                - F1 Score: A balance between precision and recall for overall performance assessment.
+                                
+                """ 
+                )
+            gr.Column():
+                gr.Markdown(" ## 🔍 Libraries Used and Their Objectives")
+                gr.Markdown(
+                """
+                ### 1. Pandas (import pandas as pd)
+
+                Objective: Data manipulation and preprocessing.
+                Justification: Used for loading, cleaning, and structuring the email dataset for analysis and model training.
+                
+                ### 2. NumPy (import numpy as np)
+                
+                Objective: Efficient numerical operations.
+                Justification: Facilitates handling large datasets and computations, such as text vectorization and matrix operations.
+                
+                ### 3. Torch & Torch-related Libraries
+                
+                import torch – Core deep learning framework for model training.
+                import torch.nn as nn – Defines deep learning model architecture.
+                import torch.optim as optim – Implements optimization algorithms.
+                import torch.nn.functional as F – Provides additional functions like activation and loss functions.
+                from torch.utils.data import Dataset, DataLoader – Handles data batching and loading for model training.
+                Justification: Essential for training and fine-tuning DistilBERT on email classification.
+                
+                ### 4. Transformers (from transformers import DistilBertTokenizer, DistilBertForSequenceClassification)
+                
+                Objective: Tokenization and model training using DistilBERT.
+                Justification: DistilBERT offers a lighter yet powerful alternative to BERT, improving efficiency while maintaining accuracy.
+                
+                ### 5. Scikit-learn (sklearn)
+                
+                Feature Extraction:
+                CountVectorizer: Converts text into a matrix of token counts.
+                TfidfVectorizer: Converts text into TF-IDF features, which measure the importance of words in documents.
+                Model Training & Evaluation:
+                MultinomialNB: Implements the Naïve Bayes classifier for a baseline model.
+                train_test_split: Splits the dataset for training and testing.
+                classification_report, accuracy_score, precision_score, recall_score, f1_score: Computes evaluation metrics.
+                Justification: Used for feature extraction, baseline modeling, and performance evaluation of different models.
+                
+                ### 6. Matplotlib & Seaborn (import matplotlib.pyplot as plt, import seaborn as sns)
+                
+                Objective: Data visualization.
+                Justification: Used to visualize word distributions, spam vs. ham comparisons, and model performance metrics.
+                
+                ### 7. Gradio (import gradio as gr)
+                
+                Objective: Building an interactive web-based demo.
+                Justification: Allows users to test the spam detection system by inputting emails and viewing real-time predictions.
+                """)
+            gr.Column():
+                gr.Markdown("## 🎉 Thanks & Acknowledgments 🎉")
+                gr.Markdown("""
+                ### 🙌 Special Thanks to Our Contributors
+        
+                **🔹 Remus**  
+                - Led **Data Collection & Preprocessing**, ensuring a clean dataset for training.  
+                - Developed the **Baseline Model**, which served as the foundation for further improvements.  
+                - Fine-tuned **BERT**, optimizing hyperparameters to enhance accuracy.  
         
+                **🔹 Ashley**  
+                - Played a key role in **Data Collection & Preprocessing**, improving dataset quality.  
+                - Successfully handled the **Deployment on Hugging Face**, making the model accessible to users.  
+                - Implemented and optimized **DistilBERT**, achieving a balance between speed and performance.  
+        
+                This project was a collaborative effort, and we appreciate the hard work put into making it a success! 🚀  
+                """)
+
+
     return interface
 
 # Launch the interface