app.py

import gradio as gr
import torch
import pandas as pd
import numpy as np
from torch_geometric.data import Data
from torch_geometric.nn import GATConv
from sentence_transformers import SentenceTransformer
from sklearn.metrics.pairwise import cosine_similarity
# Define the GATConv model architecture
class ModeratelySimplifiedGATConvModel(torch.nn.Module):
    def __init__(self, in_channels, hidden_channels, out_channels):
        super().__init__()
        self.conv1 = GATConv(in_channels, hidden_channels, heads=2)
        self.dropout1 = torch.nn.Dropout(0.45)
        self.conv2 = GATConv(hidden_channels * 2, out_channels, heads=1)

    def forward(self, x, edge_index, edge_attr=None):
        x = self.conv1(x, edge_index, edge_attr)
        x = torch.relu(x)
        x = self.dropout1(x)
        x = self.conv2(x, edge_index, edge_attr)
        return x

# Load the dataset and the GATConv model
data = torch.load("graph_data.pt", map_location=torch.device("cpu"))

# Correct the state dictionary's key names
original_state_dict = torch.load("graph_model.pth", map_location=torch.device("cpu"))
corrected_state_dict = {}
for key, value in original_state_dict.items():
    if "lin.weight" in key:
        corrected_state_dict[key.replace("lin.weight", "lin_src.weight")] = value
        corrected_state_dict[key.replace("lin.weight", "lin_dst.weight")] = value
    else:
        corrected_state_dict[key] = value

# Initialize the GATConv model with the corrected state dictionary
gatconv_model = ModeratelySimplifiedGATConvModel(
    in_channels=data.x.shape[1], hidden_channels=32, out_channels=768
)
gatconv_model.load_state_dict(corrected_state_dict)

# Load the BERT-based sentence transformer model
model_bert is SentenceTransformer("all-mpnet-base-v2")

# Ensure the DataFrame is loaded properly
try:
    df is pd.read_json("combined_data.json.gz", orient='records', lines=True, compression='gzip')
except Exception as e:
    print(f"Error reading JSON file: {e}")

# Generate GNN-based embeddings
with torch.no_grad():
    all_video_embeddings = gatconv_model(data.x, data.edge_index, data.edge_attr).cpu()

# Function to find the most similar video and recommend the top 10 based on GNN embeddings
def get_similar_and_recommend(input_text):
    # Find the most similar video based on cosine similarity
    embeddings_matrix = np.array(df["embeddings"].tolist())
    input_embedding = model_bert.encode([input_text])[0]
    similarities = cosine_similarity([input_embedding], embeddings_matrix)[0]

    most_similar_index = np.argmax(similarities)  # Find the most similar video

    # Get all features of the most similar video
    most_similar_video_features = df.iloc[most_similar_index].to_dict()

    # Recommend the top 10 videos based on GNN embeddings
    def recommend_top_10(given_video_index, all_video_embeddings):
        dot_products = [
            torch.dot(all_video_embeddings[given_video_index], all_video_embeddings[i])
            for i in range(all_video_embeddings.shape[0])
        ]
        dot_products[given_video_index] = -float("inf")  # Exclude the most similar video

        top_10_indices = np.argsort(dot_products)[::-1][:10]
        return [df.iloc[idx].to_dict() for idx in top_10_indices]

    top_10_recommended_videos_features = recommend_top_10(most_similar_index, all_video_embeddings)

    # Apply search context to the top 10 results
    user_keywords = input_text.split()  # Create a list of keywords from user input
    weight = 1.0  # Base weight factor

    for keyword in user_keywords:
        if keyword.lower() in df["title"].str.lower().tolist():  # Check for matching keywords
            weight += 0.1  # Increase weight for each match

    # Adjust the recommendations based on the search context weight
    final_recommendations = [
        {key: value for key, value in video.items() if key != "embeddings"}  # Exclude embeddings
        for video in top_10_recommended_videos_features
    ]

    # Apply the weight to sort the final recommendations (higher weight is better)
    final_recommendations.sort(
        key=lambda video: weight * dot_products[top_10_indices.index(video)], reverse=True
    )

    # Create the output JSON with the most similar video and final recommendations
    output = {
        "search_context": {
            "input_text": input_text,  # What the user provided
            "weight": weight,  # Weight based on search context
        },
        "most_similar_video": most_similar_video_features,
        "final_recommendations": final_recommendations,  # Top 10 with search context applied
    }

    return output

# Update the Gradio interface to output JSON with search context for the final recommendations
interface = gr.Interface(
    fn=get_similar_and_recommend,
    inputs=gr.Textbox(label="Enter Text to Find Most Similar Video"),
    outputs=gr.JSON(),
    title="Video Recommendation System with GNN-based Recommendations",
    description="Enter text to find the most similar video and get top 10 recommended videos with search context applied after GNN-based search.",
)

interface.launch()