app.py

import gradio as gr
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
from sklearn.metrics.pairwise import cosine_similarity
from sentence_transformers import SentenceTransformer
import pandas as pd
import re

# Load the knowledge base
encoded_df = pd.read_csv('encoded_df.csv').drop(columns=['Unnamed: 0'])

# Initialize the Sentence Transformer model
model = SentenceTransformer('all-MiniLM-L6-v2')

# Function to preprocess text
def preprocess_text(text):
    text = text.lower()  # Lowercase
    text = re.sub(r'[^a-zA-Z\s]', '', text)  # Remove special characters
    return text

# Function to generate graphs for stories with similarity > 0.8
def generate_graphs(new_story):
    # Preprocess the new story
    new_story = preprocess_text(new_story)

    # Encode the new story
    new_story_vector = model.encode([new_story])[0]

    # Calculate similarity with knowledge base stories
    knowledge_base_vectors = encoded_df.iloc[:, :-7].values  # Exclude 'likesCount'
    similarities = cosine_similarity([new_story_vector], knowledge_base_vectors)[0]

    # Filter indices with similarity > 0.8
    similar_indexes = np.where(similarities > 0.8)[0]

    if len(similar_indexes) == 0:
        return None, "No stories have a similarity > 0.85."

    # Get likesCount for stories with similarity > 0.8
    likes_distribution = encoded_df.iloc[similar_indexes]['likesCount'].values
    story_labels = [f"Story {i+1}" for i in similar_indexes]

    # Plot similarity distribution for all similar stories
    plt.figure(figsize=(10, 6))
    sns.kdeplot(new_story_vector, shade=False, label="New Story", color='blue', linewidth=2)
    for idx in similar_indexes:
        most_similar_vector = encoded_df.iloc[idx, :-7].values
        sns.kdeplot(most_similar_vector, shade=False, label=f"Story {idx+1}", alpha=0.5)
    plt.title("Similarity Distribution: New Story vs Similar Stories", fontsize=14)
    plt.xlabel("Vector Values", fontsize=12)
    plt.ylabel("Density", fontsize=12)
    plt.legend(title="Stories")
    sim_dist_plot = plt.gcf()

    # Create a bar graph for likes distribution
    plt.figure(figsize=(10, 6))
    sns.barplot(x=story_labels, y=likes_distribution, palette="viridis")
    plt.title("LikesCount Distribution for Similar Stories", fontsize=14)
    plt.xlabel("Story Index (Similarity > 0.8)", fontsize=12)
    plt.ylabel("LikesCount", fontsize=12)
    plt.xticks(rotation=90)
    likes_dist_plot = plt.gcf()

    return sim_dist_plot, likes_dist_plot

# Gradio interface
def gradio_interface(new_story):
    sim_dist_plot, likes_dist_plot = generate_graphs(new_story)
    if sim_dist_plot is None:
        return "No stories have a similarity > 0.8.", None
    return sim_dist_plot, likes_dist_plot

# Create the Gradio interface
iface = gr.Interface(
    fn=gradio_interface,
    inputs=gr.Textbox(label="Enter a story", lines=10, placeholder="Enter the story here..."),
    outputs=[gr.Plot(label="Similarity Distribution"), gr.Plot(label="Likes Distribution")],
    title="Story Similarity and Likes Analysis",
    description="Enter a new story to compare with the knowledge base. "
                "View similarity distributions and likes of stories with similarity > 0.8."
)

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